KALIBR. Russian tube tester L3-3 from 1986

Last Update: 22-Mär-2019 9:58

Link to manufacturer

The brand is believed to be "KALIBR"

On this page, I want to pay some more attention to this very nice tester. It has many advantages of a practical kind. For instance, it is very good available. Another advantage is also the good condition they usually have. Internal components are all top class military quality, and seem to survived storage and use. Once you have found the way how to operate it, and how to make cards as you need, you will see this is tester is opening new world of tube testing. The thing does what is supposed to do, with an amazing accuracy, and I can say there is no such tester available which can do all of that, for what you pay for an L3-3. .


In the internet, I have seen cards, made by using a graphics programs like SPLAN. I damaged a NOS RGN1064 with it, and I guess I was lucky the tester survived the fireworks. Also there is a card printing program around, which by itself is very nice, but unfortunately like the author writes, the program can and will have failures. Well and it has so indeed. I damaged the Transconductance circuit of one of my NOS L3-3 with it. Arghh&%$%! So I am cured from free cards. Though it worked for several tubes initially, this was not worth it.

Some information junkies are spreading this still in the internet. Well I guess I was one of those myself. The problem is, such errors can damage your tester really unexpected, and there is no safe way to test a new card. When such a card has an error, it is really possible you switch on the tester, and the same moment something begins to smoke. If that is a resistor (it happens) you can often replace it conveniently, IF you find the resistor. If you find it, the printed value is often burned off. However if the smoke comes from the cable tree, or a short underneath the plugs deck, it becomes hard to fix. If you short the rectifier tubes high voltage winding (1100 Volts AC with center tap....) I think the transformer is gone within seconds, as the transformer itself can do 400..500 Watt, and it gets all absorbed in one thin wire winding. So better don't use such cards.

The cards we sell for a few Euro each, are more like a help to the community, and all together this cost me more money as it makes. Even the punch tool was expensive. Pleas don't misunderstand me, I recommend against the free cards in the internet because I learned myself the hard way they are quite a risk to use.


Real NOS means stocked ever since and never used. In case you can buy such a tester, be aware such long storage is not good for components. Here is what to take care off.

  1. High Voltage power supply caps need to be formatted ALLWAYS after storage of more than 5...10 years. You can discuss about 5 or 10 years, but not about 30 years. That NEEDS a formatting procedure. If you don't do that, you will violently format the capacitors. This will give burn spots inside, and these are the beginning of later problems. So then you end up with this ugly blue replacements caps inside. Original caps are top quality. For formatting, you need to solder off one wire, and slowly charge them via a high resistor Like 100k or 1 Meg. You will see then voltage go up quite slowly. Same as when charging a battery. Like a 350V Cap will accept only 80Volt, and then rise slowly while you watch it . So you 80V...82... 84... 86V Etc until the maximum as written on the cap. Then you can go 25V above maximum, and they're fine. After this, like when you re try the day, you will see they accept immediately 350V. This is now a formatted cap. This effect stays for many years in storage. Once you saw this, you will understand what it hurts them to hit them with 350V right away after 30 years. If you find they are all formatted already, the machine was switched on already by someone. So hopefully professionally formatted, but better not count on it. If a capacitor has something like a voltage limiting effect BELOW the voltage that is printed on it, and you can not format it any higher, the cap is defective. Please google a little bit for this item. Some make a art from it, to repair bad caps this way. That is a bit extreme. I would rather replace a bad cap. Only the other extreme way is fully ignore this subject, just switch on the tester after 30 years, hurt all caps, and later not understand what causes specific problems. Use a Variac instead of professional formatting is better than simply switch it on. Still keep in mind, 30 years is a long time. Better format the correct way, and the old caps can all stay in and they all stay good for some more decades.
  2. After all HV caps are formatted, you should leave the tester switched on without card for 24 hours, and watch it. So not leave it on over night. You can never say what it will do. Then after 24 run hours, use some card for a normal tube, but insert no tube and leave it on like that for 1..2 weeks. This is a long time, but this is actually recommended by the factory I was told. This will internally dry the transformer, and all foil caps too. After this time is over, you will find many things stabile, that were instable before.
  3. Some maintenance is a "must". Like replace the Grid Voltage Pot meters, these are terrible quality. Look here at improvements for other recommendations.


Order Nr: 380-060-38 Anti Static treatment of meter glass
Order Nr: 380-052-26 Instruction Card: Leakage
Order Nr: 380-059-91 New Meter Scale for L3-3

Order Nr: 380-001-37 L3-3 Card for 2A3
Order Nr: 380-039-15 L3-3 Card for 5U4G
Order Nr: 380-056-95 L3-3 Card for 5Z3
Order Nr: 380-002-42 L3-3 Card for 6C45

FULL LIST: Click here

INQUIRE FOR OTHER CARDS IF YOU NEED. Almost every tube is possible.

For frequent users we sell hard plastic cards, which are offering a bit more than the original cards. Also I have nice instruction card to check the heater cathode leakage, for in case you don't know how to do that by the Russian manual. Actually I made it for myself originally :) Almost every card is possible for L3-3, provided it has the socket. Even when the socket is missing, we can make the card for the octal socket, and you make your own adapter socket. Empty Octal bases we have too (So the part that fits into the tube tester)

For nicest results, I use a plastic laminator, and hole cutting tool.


A wooden box of perfect quality can be bought from: www.office-discount.de. It costs appr 32€ there, worth every cent. You should order the metal separator set along with it. This is really a must have! Or check ebay.de for the word: karteikasten, you may find a used one.


Honestly.... The next thing missing is an English short form manual, some solution to get English text on the deck plate.

A three scale panel meter. For this, send in your meter, and you receive it back with a triple scale, in colors. (Read below for details)

Order nr: 624-024-09

L3-3 Tube tester Forum

(Click the image on the left)

This is one of the finest reports I have ever seen, about an L3-3. Written by Marc Michalzik. It is all in German language. From a simple question about some information, developed an email friendship. Marc is someone who definitely wants to go to the bottom of things. When the explanation can't be found, where others stop, this is where Marc begins. So many, many of the unanswered questions, issues, and ideas what can be done with an L3-3, are either answered here in detail, or it is described what would be the way to go. This reports sort of opens almost the design log book of the Russian engineers. Marc's conclusion is, these people really knew what they were doing. Much recommended is also his edited schematic , with all functions penciled inside. This report is definitely a "must read" when you already have an L3, or planning to buy one.

Perhaps the Google translator can one day (or already now?) view this pdf document in your own language .Download report here.

From Anton van den Oever, from the Netherlands I received many pictures fo his modified L3-3 tester. He has done many remarkable changes and improvements. One of those is shown here. You can take the Pentode for the Ug2 Voltage, and add it in parallel to the Anode regulator tubes. This boosts the power supply to 150mA. For Ug2 you need to build a new power supply, but for the few mA you need, this can be done easily with a new, small circuit. Also the problem of how to add a 10 turns pot meter to Ua, was solved here.

Ratings, at "Conclusion" Part: See here

EBAYPROBLEMS. Here we go again.

Controls in English

Testing SQ EL84

Just for Check up

Just for Check up

HQ Schematic
(Loose paper)

NEW MADE Schematic with added explanations.
Schematic with kind permission from Marc Michalzik. THANK YOU MARC!

Schematic by Andreas
(see L3-2 Forum)

Drawing with kind permission from Marc Michalzik. THANK YOU MARC!

Some drawing,
from manual
(the cable tree)

Calibration Report

Calibration Report
random page

How it works L3-3

Ground can be
detached this way

KALIBR. Model L3-3.

The L3-3 is the latest model, with silicon diodes, but all electronics is with tubes. The L3-2 is similar, but has still tube rectifiers inside. The L3-3 is is one of the most amazing testers I know. When I bought mine in 2002, the tube tester market was not so over hyped as it is now, and this tester was regarded an unknown Russian product of little interest. Right now, things have changed. Understanding of this product has improved at least 15% over the last ten years. (Yes, this means what you think it means:)

Detailed picture. I have two of those, so I can measure large numbers of tubes faster. While checking the readings on the one tester, the other one will warm up the next tube already. Like this I can work really quick.


A small introduction to begin with, the L3-3 is one of the finest, or probably THE finest vintage analog tube testers there is. The circuits design is extremely mature. Apart from the 10% type capacitors, and eletrolytics in there, each single component has it's own quality grade, suited for long term problem free functioning, up to today. I found only one low quality part so far (see below under improvements) and it's not so much a bad part, but it was not over specified. The 5% caps and mica caps seem all still fine. The accuracy of this tester, is not coming from precise calibration, but from precise passive components inside. The most difficult part, where many tube testers produce rubbish, is the transconductance measurement. Here, the KALIBR has the option of a quick user's calibration before doing a measurement. A precisely known reference signal, which is produced by the machine itself, is compared with the equipment's own result, and the machine will now correct it's own error. After this is done, the tube is measured. This method is still found in modern digital multi meters that are self-calibrating, like the Agilent 3458A multi meters, which cost the same as a new car. It is missing in the expensive Amplitrex AT1000 digital tester, but ok that would be too much to ask for what it cost. But.... hey... the L3-3 has it :) This brings the precision in the range of less than 1%, provided the whole tester is mint condition. Even so, with a lousy calibration you will highest precision, since the reference they use is only made of precision resistors. Brilliant as it is, this method even compensates the error of the panel meter. The only thing you achieve by calibration is stability. So it would be a misunderstanding that you get more precise results by an internal calibration. It is just that the with a bad INTERNAL calibration, the EXTERNAL calibration (on the front deck) needs to be done before each measurement, and even so the nano ampere measurements may become hard to do because it drifts away while testing. With a good internal calibration, the EXTERNAL calibration is done only once, and then you can measure whole series of tubes with it with a beautiful precision which is identical to the panel meter's reading. So what you read is then guaranteed what you have.

1 The tester has a tube regulated 250V supply voltage, which is used for the internal oscillator, the band pass filter of the transconductance test, and for the electronic nano ampere meter. Each of the three depend do not depend not so much with their function on a precise 250V, but they depend with their stability on this voltage to be THE SAME as always. The oscillator frequency, depends very critically on this 250V. The closer you are to the center of the band pass filter, the more stabile the transconductance test gets. So not more precise, but more stabile. This means you don't need to do the EXTERNAL calibration very often. So adjusting the 250V, is one of the first things that helps here. After the tester is warmed up, it must be 250V, but mainly it must be STABILE. So if it drifts, you have a hardware problem.
2 Adjust the mains voltage calibration. This is a more difficult one, but there are some reasons why this calibration is off specs sometimes very much. . (More information here)



Features. It is a fully programmable tester, by sticking pins in the pin board. All voltages are internally stabilized, Max 300V, 100mA for the plate, and 2x 10mA for the other adjustable power supplies.   It features an internal 1400 Hz oscillator. For leakage tests there is a tube microampere tester with 2nA resolution which you can calibrate from the tube deck easily and quickly. (Try that with your digital multi meter - forget it). For perfect leakage of heater to cathode voltage can be programmed. Also the internal tube screen that separates double-triodes like some ECC....types, can be tested for leakage. The leakage current can be measured very precise, and also the voltage can be checked at what the leakage will start to begin. Tubes filaments can be heated with fixed voltage AC or fine adjustable DC.  Like in real amplifiers! 

A very special feature of this instrument is this: You can program a Single Ended circuit for the tube under test, choose the bias resistor and the bypass capacitor. This is not the only circuit diagram feature. With an internal 1400Hz (distortion free) oscillator you can test what the tube does, and make AC measurements. Wow! Can you imagine what a machine this is? It took me 3 weeks to understand most of it. It is constructed with printed circuit boards. Most of the electronics is with tubes, and some solid state components. This is the most versatile, and most precise tester I have seen so far.  I was lucky to buy one 1986 model NOS. I am the first user of it. Most calibrations can be done on the front panel, using the front panel meter. It can work on any mains voltage from 100...380Volt, and even 400Hz mains as used on ships and airplanes is possible. There is nothing you can NOT do with it. 

The functionality is so excellent. There is simply NOTHING you can not measure, apart from heater current. The rest, you name it, the tester can do it. The transconductance measurement is done in such a way, that even curves unlinearity is eliminated by filtering out the harmonics from in the measured AC plate signal. For this a very sharp band pass filter is used. Same method as used with classical distortion measurement instruments. It would not surprise me, if this is the only tester ever made, using such fine electronics for this function.

Improvements and some repairs

Well, some can always be made.

Banana Plugs. This is the only repair, you MUST do on any L3-3 right away, and not wait until you damage something. In the tester deck, you have five banana plugs. Two for the Anode, two for the grid, and one for ground. These are definitely unreliable, and should be replaced with any tester. They can easily cause damage with the tubes, or with the tester itself. Even so, the L3-3 that needed this repair, I bought it with a (nicely hidden...) cable tree burn damage. So it is possible, this defect caused the cable burn in the first place. All I can say, spend 2x 50 cents for this repair, and prevent big trouble. Read more here.

Isolate the mains switches. Just very practical. People say when you had many electroshocks, you get less sensitive. I can confirm this is so. The best one I had was 850 Volts, just from an amplifier, I fell on the ground and I could not work for three days. I don't know if it has something to do with it, but touching 250V by mistake doesn't seem to bother me. I just say ouch, and I continue work. Of course that is imaginary, it's just how you experience it, since in the end a certain DC current will be lethal anyway, and I believe that is appr 60mA from one arm, and an unlucky path though the chest. Yet I had shocks just through my hand, with burned spots, and small blisters the next day, and when I happened it did not feel very serious. I just continued what I was doing. Only problem still, I don't like it at all. So when working on the L3-3, there are a few places with mains voltage on it, even when switched off. It just feels better when they are isolated. Also one of big the capacitors. That's a bad place to touch, they're large surface, and all in the front, waiting for your fingers. These pictures here speak for itself.


Order nr:

Three Scale faceplate for your L3-3! The meter has only one scale (0...150) but this also used for 0...75 and 0...30. This is such a pain, and all people use a calculator with it. I made a new meter scale for it, with all three ranges. This improves the user comfort so much. The test cards we sell, can be used with the normal "150" meter of course, but they become even more comfortable if you use the new meter scale. Reason is, settings are difficult sometimes when using others than the 150 scale. For instance how to set 47Volts at 75V Full Scale needs a calculator. You need to calculate: 150/75 =2 and then 47 * 2 = 94. It gets even harder when you want to read something, because you are constantly calculating. So when you reed 92.5 on the 150 scale, it means 46.3. Quite nasty. And then you read 93,7 meaning 46,9, and 88,9 meaning 44,5 and... actually this drives me crazy.

Now with the new made test cards, we give you this classical setting (so 47V at the150V scale) so you can use the original meter scale if course. Also, we give the position at the "15" scale. In this example it simply means set it at 9,4 of the "15" scale and you're done. So for a maximum of four settings per tube, that saves a lot of trouble, and also you can make no mistakes. Additional there is a red line at "6.3", and if possible always the heater voltage is on the 15V scale. So that makes it comfortable and reliable to set any 6.3V heater tubes.

Art work / re work the meter scale

Note, with a factory production run of high precision meters, the scale is printed such that it matches the natural unlinearity of the actual movement inside. Then, at final assembly, the meter itself is adapted to point exactly at the right position of the scale, this can be done with the balancing weights on the coil, and some other tricks they have, for changing linearity of the movement system. So this is for linearity, not for full scale. (As full scale is easier to adapt)

In the end, movement system and face plate are a pair. Only like this you get a precision meters. So when we make a three scale face plate, it is not possible to make a universal face plate for all meters, because we we must preserve the original divisions. The unwanted texts are removed carefully, and then the faceplate is repaired with transparent paint. Over this is glued a transparent foil with the triple scale, but only the numbers. You won't recognize this as a foil. Appearance is perfect, as if it was made like this originally.

As a final step, the meter glass and inside gets a new antistatic coating, as this is often gone with older meters. Also the needle is re coated, because otherwise the needle paint or plastic inside parts can charge up statically, causing remarkable errors when you polish the glass. Some people report antistatic charge to remain inside for months, or in very dry air, the needle paint can re charge just by needle movement. So re new the coating is maintenance, and is included.

We l add a fine red hair line at 6.3 Volts at the 15V scale. This is very convenient to adjust the heater of 6.3V tubes. At the "Calibration" point is additionally the work "cal" now, and a small triangle. So you can't miss the calibration mark any more.

Price is 85 Euro + 19% VAT for Europe + Shipment.

Some hints:

  1. If you want to go for this, short the meter from behind before shipment. This protects the movement from shocks. The L3-3 meter is very good quality, but still please use a large enough shipment box.
  2. In case some defects appear when we open up the meter, we inform you about it first.

Reverse the "test" button. To see a test result, you need to press the "test" button always. If not pressed it shows the heater voltage. I don't know if you have an iron man thumb, but mine starts to hurt after testing more than 20 tubes. With a careful modification of the "test" button switch, these functions can be reversed, and the user's comfort increases very much. After the modification the tester indicates the test values by default, and when you press the button, it shows the heater voltage. Read more....

Check this, if the main "test" button under the panel meter seems to have dirty contacts. It probably has no dirty contacts at all. I have been struggling with contact errors with the "test" button switch of one of my L3-3. Contact spray was helping sometimes, but it got worse over time. So I decided the switch had to be taken apart Well first thing that became visible was the defective cover. I did not expect that! This cause all that trouble since years. The contacts could not move freely inside. I repaired it with hot air. I also took the chance to reverse the switch for this L3-3 too (see above). Large Picture
Change the Ug1 (grid) potentiometers to a 10 turns type. Now, adjustment is done much easier, and also the temperature drift with the Ug1 is gone. I would call this a "must" for all L3-3. You may not have expected this, but temperature drift of the original potmeters is really terrible. The improvement is stability is very large with a wire wound pot meter.
Change the Transconductance Calibration potentiometers to a 10 turns type. This adjustment you will also find kind of instable, and you need to repeat it almost before every measurement. There is more than one reasons for this drift, and it is 100% sure, we have not found all of them. One of then however is the original carbon potentiometer, which has temperature drift, and also the regulation is much to imprecise with a one-turn potentiometer So this change will greatly improve the working comfort.More about it here.
    This is about how you MAY improve the combination Audio Oscillator and the Bandwidth filter. These two cooperate, to do the transconductance measurement. The bandwidth filter is very sharp. It is extremely important to understand, for a precision transconductance (Gm) measurement you need to reject from the measured signal, the TUBE DISTORTION which you always have since tube curves are no straight lines. Also should also be clear, that any power supply hum may also not be measured together with the real test signal. Also hum radiation from the outside world into the tube, will exist and must be rejected in some way. This may seem very logical to you, but it is the major ERROR SOURCE of almost any tube tester, including the expensive digital testers, that ignore this problem as well. So yes the L3-3 may come sometimes with a little less Gm as you may have expected, but now you know why that is! So the oscillator produces a signal to begin with, which is a "nice" signal but nothing special. However the Bandwidth filter is special. It is extremely sharp, and it reacts to 1 few Hz changes already. Before a measurement, you need to adjust the oscillator, by connecting it to the bandwidth filter, and set the needle to the calibration mark. By this the sensitivity of the filter is adjusted to the base frequency (fo). So all distortion components of higher order (by Fourier) are f1, f2, f3, etc, these are all rejected. Also the mix components of the mains hum are rejected. That is 50Hz, 100Hz, but also f0+50Hz, f0+100Hz, f0+150Hz. The whole junk of the oscillator gets filtered out. Now comes the transconductance measurement, by the flip of a switch, and again all junk of the oscillator gets filtered out, but also the distortion products of the tube itself. So you really talk about Gm as pure as can be here. To my opinion, this represents the only correct was to measure transconductance and nothing else. Understanding this, will also explain why static and dynamical measurement will give a difference that way. Also when comparing the normal (dynamic) measurement with a hand operated static al measurement done with the L3-3 and a digital meter. Now comes the improvement. Or rather what it should be. The audio oscillator has a small drift, and whatever the explanation may be, it's a mystery. After thermal stability, let's say after one hour, the oscillator still drifts further, and it keeps on drifting after one day, and even after three days. Now drift is very small, and you can easily adjust it before a measurement. However this is annoying, and no explanation was found so far. Please note, by itself the drift is amazingly small for a tube oscillator Perhaps a few Hz per day only. However the Bandwidth filter is so sharp, it doesn't like this. My question is, WHY is there no thermal stability, also not after three days. All L3-3 seem to do this, it is no defect, it is just how they are. So you must always set the meter to the calibration point before you begin, but this lasts longest 30 minutes, and you must re-calibrate. Also a complete overhaul of the Audio board, as done by Marc Michalzik brought only marginal improvement. This is really a tough matter. Do not see this as an error! I see it as something which MAY be improved. If you know how to fix the oscillator within a few single Hz, you can let me know. More here
    Plate voltage, it is also a good idea to replace the voltage pot meters by a 10 turns meter, however that needs digging into the circuit, because the required 1Mega ohm 10 turns pot meter doesn't exist. (on earth)
Some small progress with this now. Add PC interface.

Add Socket for Au8 socket tubes. This socket (AU8) is missing painfully. A huge variety of Audio tubes was designed 1930's until 1950 by Philips, TFK and Valvo. The original socket #2 is a type for some military tubes, that are not interesting to me. So I decided to remove that socket and make room for the Au8 socket. After removing the original socket, the nice surprize was, the mounting hold and screw holes EXACTLY fit a standard NOS AU8socket, and wiring is simple too. The result looks as if it was all original intended this way.

More here

An operation of the Switch board.

This tester was born with a defect in the cable tree. At the age of 39 years, the tester had a brain operation, to cure his malfunction.

Report here.

    Add banana jacks to measure all voltages and current. Project by Marc Michalzik.

Grounding. This is an interesting issue, because we ground equipment for our safety. For a tube tester, the question is what is safe. First look at what happens if the tester is not grounded. With a high quality mains transformer, any contact risk of the user with the mains voltage is small, and could only come from defective internal wiring perhaps. What is more realistic, you touch the +250Volt. Suppose you touch a plate cap, then you get only an electro shock when you touch the L3-3 case with your other hand. Since you do not ALWAYS touch the case with your left hand, the hazard is reduced.

Now for the case you ground the tester, things change. Now, the current path trough your body is also possible when you touch any OTHER random grounded subject, or when your body is grounded whatever other way. Like via a grounded workbench that you lean on. What is eliminated is risk of mains voltage contact. So that is the advantage.

So often I hear people say you" must always ground" a tube tester, because otherwise it is dangerous. Is that really so? Well I would agree for a 60 years old Tester that you buy on Ebay, and which has been rotting in somebodies garage for the last 25 years. For a a NOS L3-3 it is another situation. I do not give a "best" advise here for the one or the other method, because each must choose for himself what is best. I have made a mains cable with a separate ground plug. So I can choose as I want, and I must say for myself, the risk of an internal wire problem, is small. Myself though, I touched the HV by mistake often though. Often I use the L3-3 ungrounded and then not much happens when I touch a plate cap. But again, this is no advise to anyone. It is only what I do myself.

Lifetime and stability. I have one NOS, and three used. With the NOS, the Grid1 voltage needed no more adjustment after one hour. It stays stabile or the rest of the day, whereas the used one needed hand corrections. I mean here adjustments in a few 1% range, as fine as the meter can indicate. So you see again here, NOS has some advantages. The tubes in both still the first, original factory tubes. I have about 5000 on both testers now, and all original tubes still inside. I can use one tester to test the tubes of the other. Later, I found the drift s was caused by a temperature coefficient of the G1 pot meter. I replaced them by a 10 turns heli coil, and Ug1 is more stabile after 10 minutes, and stays like that all day.

Maintenance. Who used old instruments, containing power electronics, will all know they need a repair sometimes. Something breaks here, something gets s defective there, and it just belongs to daily use of an old instruments. One L3-3 I have is NOS, means it was never used before. That is not always an advantage for old equipment, because often that means often all capacitors are bad. The other one was used very often, and it was officially calibrated every year until 1999, which was the last calibration sticker on it. The NOS one stabilized perfectly after 100 hours, and the used one worked perfect right out of the shipment box. All I had with mine, was a minor problem with a plastic cover from the inside detoriating. That I can replace easily by a wooden cover. My L3 never needed never any repairs at all. Only service, like a defective control lamp. All parts are still original, even the tubes. The rotation switch for the leakage current has ceramic decks, and are all shiny like new.

In case you ever do need to repair it, you need to know this: First, you probably won't be able to read the Russian manual. Second, you don't need it at all. The schematic is very well drawn. They didn't try to to make "clever" schematics, but one that normal people (like me?) can understand, and it's all on one huge sheet of paper (size of 80cm). The tester can open from all sides by means of perfect mechanical systems. Each wire is numbered, and the number is on each wire with a small sleeve. Third, the electronics function has no patented crazy circuits that nobody understands, like the Hickok testers. Instead you will find straight forward electronics, nicely arranged. It has printed circuits for the oscillator and the regulated power supplies, and free wiring for all "heavy" stuff. It is all made, as I would make it for myself.

Schematics. The paper work that is around, seems to have some versions, but unfortunately the drawings do not tell nicely for what version or series numbers testers they are. I will try to list up some characteristics of the L3 Versions, with the schematics belonging to it, but now it is incomplete. I hope you like these pages, and you can do something back for me, by sending information you have and share it with others. Thank you!

Meter circuit like this, with R43.
Look here and here
Schematic NEEDED
Early schematic version, that was with one of my testers. It is the version with regulator tubes, instead of Zener Diodes.
Version By Marc Michalzik, with all values and functions in it. This took weeks to make it .Thank you VERY much, Marc!
This R43 is to calibrate the Panel Meter

A Russian Data book, with all tubes + Semiconductors. For instance the Bulky DIodes are in Page 128. You need the djvu viewer for this.
Book is here


Interesting application: Since the L3-3 is all DC regulated, it is also a universal power supply for building new tube circuits. It offers a range of bias resistors, that you can choose from. You can AC or DC heat the tube you are working on. You have a stabilized anode voltage, witch DC current measurement on it. You will find, if you do not draw full 100mA current, the stabilized plate voltage is a lot higher than specified. That is because the specification is at full load. There are banana connectors for: Plate, G1, and G2. If you know how to drill a hole, you can add three holes. One the "Ground" and two for the heater, and there you are: Your universal DC regulated tube power supply. Of course with current measurement on any of the connections. When testing for transconductance, even a 1400Hz test signal is applied to the grid. Isn't that nice? Mind you, this needs the making of a special card, which sets the voltages and current ranges right.

Some notes, if you want to buy one.


EBAYPROBLEMS. Here we go again.

2) Model. I do not recommend to buy any other model than L3-3. Models like L3-1, L1 etc are a lot older, have tube rectifiers inside, the electronics get much warmer, components inside are 20 years older, and have seen a lot more use. Also such a tester probably is not NOS. The electronic components of the L3-3 are from the late 80's and good material was available at that time. I can not say this from the 1960's. There is a totally messy German translation around, but here is a very nice, better version:

3) Language. Too bad, it is all in Russian. WIth a little fantasy you can understand what is written on the tube deck and the cards. When you are not stupid you will get around with it. Like "mka" is micro Ampere.

German Handbook (Improved version)

Russian Handbook

English translation (much unfinished)
Google translation is making enormous improvements It is impressive! If some texts look strange, try again in 6 months.

4) Price. The market price of a good L3-3 is based on the fact that somewhere some people have an NOS stock still, and also the tester is a bit of a secret even now. However both things will change. So one day many will know this tester, and NOS stock will gone. It is the ever repeating NOS tune, like an LP record with a crack in it. When it's available all want to "wait". And when they are gone, they fight for second choice ones. This tester is many times better than the AVO Mk4. I once saw people pay a record price of 3662 Euro for a MK4, in 2010 (Auction Nr 370415230702). What a crazy world we live in. At that time you could buy a whole pile of NOS L3-3 for the same money. But I guess it's normal. It's the world of NOS materials.

5) What you get. I do not know how many of those testers are on the market, but I see several people selling real NOS one. A nice, all complete, NOS is the best buy, though a clean used one is probably just as good. For a NOS the price adder is appr 30%. If complete, they are equipped like this:

  • Green wooden army crate. Inside is:
  • The tester. Check if the card plugs are all with it. They are in the tester front panel, in blind holes
  • Mains cable, with it's special chassis connector.
  • Wooden accessory box , containing:
    • Card set. Make sure the universal card is not missing.
    • Spate tubes, complete set
    • Small items, like plate cap cables, etc
  • Paperwork three pieces, each with series number of tester on it:
    • Handbook, with full construction details
    • Original Schematic, printed on large sheet
    • Calibration booklet.
  • Cards. If the card set is missing, that is a real pity. The original cards for tubes like 6922 and 6SN7 are masterpieces of "card design" and show you how to get the utmost out of your tester. (Russian designation for those tubes are another though. For instance 6SN7 is called 6H8C). Making new cards for tubes that have no original card, is DIFFICULT. This means you need to understand the tester very well. I have written a software for making cards, so it prints cards in color even. However there is almost no tube it can not test.


Ebay is often the only way to get an L3-3. Please keep the following in mind:

a) Defective ones and buying risk in general. Sometimes you see them offered as unknown condition. Remember, tube tester trouble is general is extremely difficult. Often you find the defect is a burned transformer or a damaged meter. Still, the L3-3 is much easier to repair as any other. Electronics are normal, and accessibility is a dream. The chassis has wings that you can open, and even each wire has a little plastic number tag on it, which corresponds to the schematic. In any case, finding replacement parts may be possible when you are patient. Still the difference in trouble between a really good one, and one with problems is HUGE. However price differences for such are small. So better take a real good one, and pay a little more.

b) Shipment damage. Let's say it is normal, the aluminum case will deform the bottom feet in shipment when the box gets dumped. A tube tester is not a bag of old apples, but you can try to explain that to UPS. (Youtube for "UPS throwing"). If this damage happens, just take off the complete case with only four screws, and you will find all inside parts are fine. Just the aluminum case bottom is deformed. It is very easy to get the case bottom perfectly in shape again with a wooden hammer. You won't see it any more.

c) Humidity. I do not know the reason, but many of the NOS L3-3 have been stored humid, and the inside will show some white oxide here and there. However humidity does not seem to do any harm to the L3-3. Any part that would suffer, is well protected. Such as completely lacquered printed circuit boards, and the leakage switch being all ceramic.

These can be very good, even though the inside piece parts seem to have an oxide layer here and there. As far as I know, there are no parts inside that suffer from that. The critical part, the tube nano-ampere meter and oscillator are on one PCB that is dip-lacquered, and the leakage switch has ceramic decks. Such an NOS tester is more recommended than a heavy used one, since a used one is not much cheaper, but may have the typical issues any used tester has.

d) Payment conditions. Decide this for yourself, but let's be honest about this: If the seller wants pre payment by bank, you take all the risk . What is the problem with accepting Paypal, when the seller guarantees the tester to be good on arrival, and ships well packed with full insurance? I repeat it may be too often here, but on Ebay I see always the same two problems when selling tube testers. That is: hidden but severe problems inside the tester, and shipment damage which becomes the personal problem of the buyer. So specially for this, Paypal is the best way. And yet, you loose shipment cost on the way back, and getting back 200 Euro tax and duty is a path full of traps. I would say, that is normally not working.

Calibration procedure as used by one of our customers

CONCLUSION: Amongst the top-ten testers. Best buy ever for the money. It gets a well-deserved FIVE STARS ***** (what is this?)

prices seen:
NOS 430€ (my tester)
All complete
no cards
399. Complete. incl shipment to Europe.

"So called" NOS
All complete
785 incl shipment to Europe.