AVO Tube tester MK3/Mk4

Project 1

Digital AVO Mk4

A bit of a crazy project, but turns out very interesting. Look here.

Project 2

Just checking up this AVO turned into a full restoration.

Look here

AVO Mk4 curve tracer modification


The very remarkable function of the Ohms Meter of the AVO Mk4

Perhaps we say it was not an error, but they found no better way.

Meter availability

Original replacement meters are impossible to get. What you see on Ebay reflects that a tester without meter is worth very little. You can't buy the meters, and who knows what's wrong with a tester that someone removed the meter from. I have only once seen once an NOS meter on Ebay, ever since 20 years, ,and it went for appr 700 Euro. So chances you find a new meter of Ebay, is virtually zero. Yet the price of 700 Euro was justified if somebody puts it in his tester, and sold the bad g meter on Ebay. There is always somebody thinks he can repair it. Picture1 - Picture2 - Picture3 - Picture4

Meter Repair

This is possible in most of the cases. The only thing that can not be repaired is a dead coil, though this may only appear so, and you have only a bad contact somewhere.

A repair is always a complete overhaul at a flat rate. I can not just repair a loose glass, and leave all the dirt inside. Also, most meters I had here, have seen some repairs already, and some kind of unsolved issues, or damage by actions intended as repairs. So a complete overhaul is always needed anyway.

We have a meter for sale with adjustable sensitivity and adjustable impedance, both from the back with 20-turns potmeters. These adjustments are added to the meter. It was made from an unused meter from the 1960's, still in the box. It is adjusted for 30uA, outside the tester. For better precision, you can calibrate it inside your own the tester, so you can compensate for some influence of the iron mass of the tester. Also you can now re-calibrate the meter at any later time.

  • 30uA full scale +1/- 0.5% full scale.
  • 3250 Ohms impedance +/-1%
  • Magnetic damping, build into the meter by me. This gives better protection.
  • Old style housing.
  • For the Mk3, Red LED lighting is possible at extra cost. So you will still have the red scale light in case the protection relay activates. This requires no additional circuitry, and takes the power supply directly from the AC power supply of the red light bulb. In case you like it, you can still add this (Mk3) option to the Mk4 tester without any changes at all to your tester. The meter will get two solder terminals, which you need to connect in series with the protection relay, and then the red emergency backlighting works.
  • LED front option is possible too at extra cost.

This meter is 400 Euro without VAT tax. In Europe, 19% VAT tax applies, and the price is 476 Euro.

Backlighting in Red 100 Euro. In Europe, 19% VAT tax applies, and the price is 119 Euro.
Front lighting in white 100 Euro. In Europe, 19% VAT tax applies, and the price is 119 Euro.
Both options together is possible.

Mechanically this meter almost fits in by itself. You need to file the hole distance of the mounting plate inside the tester 6mm wider. This is really easy to do, and uncritical for precision, and mounting itself is universal, by means of the mounting plate.

Note, if you adjust the sensitivity, you need to re-adjust the impedance for 3250 Ohms +/- 50 Ohms. Though impedance change is only small, the attenuator switches of the AVO work more precise if the meter is exactly 3250 Ohms. Please add a 10uF cap in parallel to the meter if you want to measure impedance, so to protect the meter from over current.

Side view

A Simple way to test the meter of your tester

Some raw information about a sticky roller switch

Screws. The AVO Mk4 originally had a plastic dust cover with it, and it's an extremely good thing when that is still with it. Sure such an AVO was not kicked around, because the plastic tears easily. So when it's nice condition, that's a perfect sign. Not just the cover is worth 40 Euro, but I find the dust cover is a very good protection against oxide of all metal parts, and against sun bleaching of the adjustment wheels. Also the paint is a cleaner color then, there is not this brownish faint, and the slot for the tube book is cleaner also. So in many cases you have quite a good Mk4, but it has rusty screws. Now, these are not just any screws. It's obsolete British thread. In history there was some sort of a war going on, where all countries invented their own thread. In The USA they are still trying to prevent metric thread, but I give them little chances. Metric is too logical to ignore. Anyway, when equipment is older, you need to research a bit on that. For the Mk4, the front, back, and side panels are so called british B4A thread. This was a "fine" thread, once upon a time in use for clocks. Well and they used that for the Mk4. It's quite similar to American 6/40 thread, but B4 has 38 TPI (threads per inch) and 6/4 has 40TPI. For the back panel and front panel you can use the easier to buy USA fine thread 6/40, though it's also a vintage product. For the side panels you can only use B4 thread, because the screws go in quite deep. So then these 2 threads difference make a problem. You can overcome that by cut off the screw at appr 5mm. (first put a nut on it, then cut it off. Such a nut you can borrow from the inside somewhere). The front plate screws are called "slotted raised countersunk B4 machine screws". You can google for that. Some UK companies still offer those screws for acceptable prices, like 1$ each. Same as what vintage 6/40 screws go for at Ebay. The above picture is from some screws I bought on Ebay. The below picture are new made screws I bought from a German company: www.astro-electronic.de. They are 6/40 and I used them for the side panels, behind which is the fuse and calibration. Just for fun because they looked so nice. However they are 6 Euro each, incl. local shipment.

Some unofficial test points for the Mk4 only


(Some small bits of) Company History

Address: Avo Ltd, Avocet House, 92-96 Vauxhall Bridge Road, London, SW1

  • 1923 Founded.
  • 1958 Took over Taylor Electrical Instruments Ltd, UK.
  • 1959 Metal Industries Ltd acquired Avo Ltd, including Taylor.
  • 1967 Thorn Electrical Industries took over Metal Industries
  • 1979 Thorn Electrical Industries merged with EMI to form Thorn EMI.

Background of this tester.

There is a lot more to learn for me. If somebody has more information, I will be please to add it here.

As the MK4 was made, the previous model was the MK3. The differences are minor. Rather Mk4 is a face lift version of the Mk3, but when you only look at what it the functionality, I see no difference.

Differences between Mk3 and Mk4, I noticed myself:

  • Mk3 is less easy serviceable. With Mk3, the "electronics board" or whatever it is called, is mounted on the machine bottom, whereas with Mk4 it is mounted behind the left side panel, and easier accessible. So check this link for the Mk4 if you have not dome so before. With Mk3 this panel is mounted on the machine bottom, really nasty to access.
  • With Mk4, the tube rectifiers for the grid voltage are replaced by silicon. As this is really very old silicon, these Lucas Diodes are very good, with gold plated wires, and a two components epoxy, in a filled cup. Really gentle for the inside chip. Sure better as molding it, as done today for diodes. The proof is, after 50 years, they are still fine. Unlike tin plated wired, no oxide will creep in along gold plated wires. I checked those on the Tektronix 750 curve tracer, and they are simply excellent. 750V break down voltage, and no leakage. You can a non destructive break down test on the Tek 570, with l current limiting. I see no reason to replace them.
  • Mk4 has a gear on the dial wheels for Gm and Ug. That sure gives a nice feel, but not any better precision, as the potmeter underneath is a normal wire wound pot meter as any other.
  • Slightly difference arrangement of the knobs.

Mk3 and Mk4 are more sensitive with the leakage test, this is a very nice feature and places actually these two on a level never seen before with any tube tester before. As far as I know this level was never reached by any later testers as well. Test current is so extremely low, and though the source voltage is 160V or so, at only 30uA short circuit current, this is so little you do not even damage a semiconductor with it, as the voltage drops immediately on a short, or also on leakage. With Mk4, the precision of the dial wheels was improved, and some more little things. Yet overall a Mk3 is also a beautiful tester, and can hold pace with the MK4. It must be said . The weakness of all AVO is the current sense resistor, and it is just this nasty resistor causing errors with low impedance tubes. More about this error is here.

Before the Mk3, we had the Mk2, which is also a very nice tester. The Mk3 has a much stronger case, and a bit different way to operate the Gm test. Comparing the Mk2 and Mk3 I would not say one is better than the other. The Mk2 has the softer aluminum case, which can be damaged in the post easily. To me it seems, people hammer often on instruments, so case of the Mk2 often looks not nice any more. The Mk3 and Mk4 case, can be hammered on, and since people seem to do this a lot, they survive often in nicer condition that Mk2. The case of the Mk2 however is made out of one piece (nicely!) and if you want to service it, you will LOVE this option. You will not like Mk3 and Mk4 case, for servicing. Also the Mk2 had a much wider range of the grid voltage, and sure for testing some of the low gain triodes at very high plate voltage, you may find this an unreplaceable goody of the Mk2.

Also, you will see the Mk2 connects to the history, of how they came to this kind of tester. It is an improved Taylor 45C. So Taylor actually invented this testing method, and this kind of measurement of Gm. AVO as they were building the Mk1, simply bought the Taylor company. The early Taylors had only Gm measurements, and the later 45's had also DC current simulation (using AC Voltage - same as all AVO's). Along with TAYLOR there was the French company CENTRAD, offering something amazingly similar. So these are the roots of the AVO tube tester.

Going back further in history, we find the Mk1, and here the path separates. This is really an original AVO, so to say. It can only do Gm, but wow...! It will do this so nicely. I have one here, and I can tell you it is an amazing piece of work. Precision is stunning, and the good part is, there are no calibration points inside. Mk1 can do ANY socket connection, as it already hat that marvelous AVO switch. The AVO tube data book was updated for the Mk1 way into modern times, so you have lots of data for modern tubes, AND there are blank sockets positions, so you can mount a few of the modern sockets in there. So when data is missing, take a tube data book, and do it yourself. So Mk1 basically can still today do ANY tube whatsoever.

Mk4 Weaknesses

I start with weaknesses, because the rest is all strengths. It has it's weaknesses, but not many. These are:

The meter and the calibration. Though these testers still do a basic job with a not so nice meter, and out of calibration, the way of functioning in that case is only telling if a tube is good or bad. Owners of such an AVO are shocked to find out they have been using the instrument for a decade in that condition, thinking they had a "reference" all those years. Reason is, they have no calibration tubes to check the tester, and when they do get some and the tester indicates those tubes wrong, they begin a dispute about the calibration tubes. It takes them often a while to understand, their AVO is out of specs.

There are two kinds of calibration. The one is "maintenance" calibration, also called user calibration. Though the word user calibration is misleading, because only very experienced repair men are capable of doing so. The other is factory calibration, which is only needed when the pot meters from the dial wheels are exchanged. So you have these three things: The meter, maintenance calibration and factory calibration. A fourth factor would be: No defect parts inside. These things all interact, and when one of them is not good, there is no point in trying to fix that by the others. If you do, the tester gets fuzzy. With that I mean, some tubes are indicated correct, others are not. Or plate current is indicated correct for some tubes, and transconductance is indicated wrong. Or vice versa, and it is simply a mess. It is extremely important to understand the strategy to get an AVO in 100% good condition. First, all internal parts must be good. High resistance values, with low tolerance on them (read the manual's parts list...) are suspicious. So are carbon resistors. Capacitors are very few inside. The damping capacitor across the meter is always bad. Next factor is: The meter must be fully within specs. A sensitivity error, or linearity error, would mean you calibrate the mains setting wrong before you begin. Then, when these two factors are perfect, you can do the maintenance calibration. Then you need to verify if some nut didn't mess with the factory calibration, and then comes the maintenance calibration, for which you need two or three reference tubes.

Under normal conditions, calibration actually is seldom needed. However in reality the calibration is often messed up by somebody. Rather the meter is always bad in some sort of a way, and the user calibration is simply misused to compensate for that. This way to go is not working well, and when you have bad luck, some previous owner went a step further, and also upset the factory calibration, in a hopeless attempt to make things better. Then, a calibration without good calibration tubes, it's just a joke, but people do it. (Sorry for the brain wash).

One more weakness, a little bit out of the context here, is the serial resistor in the anode circuit. This resistor is 199 Ohms, an it is simply a bit high value, but I understand it, there was no other way. First, this makes measurements a bit imprecise for tubes with low plate impedance, but there are ways to test such tubes still. Second, this resistor burns down when you test a 5U4G rectifier and let it on the tester for longer than 5 minutes.

The AVO MK3 / mk4 meter was a masterpiece 50 years ago, and it still is. When you want to buy a replacement, you will see these are not for sale anywhere. This extreme sensitivity (30uA) is not supplied at this low coil impedance (3250Ohm), and at this needle length. (We can supply you new build meters though, that we adapt to be electrically equivalent to the original)

The original Mk4 meter has no internal damping and the damping comes from an external capacitor, which is electrolytic, and is always dried up. Generally the old meter is not resistant against operator errors, and precautions inside the tester are not sufficient. Probably all AVO users had sometimes this situation to set a switch wrong, or whatever it was, and the needle was hammered in the corner. Most believed the meter survived it, because they see no problem other then having to re set the zero screw. But did the meter really survive? Probably not. This 30uA coil is so fragile, it deforms by overload. Good linearity is then gone, and full scale sensitivity is reduced. Have you tested the meter linearity afterwards? It can have 15% linearity error due to coil deformation. Things get worse when such a tester is re calibrated, with the damaged meter inside. In the past 50 years, most of these testers have seen quite a history of owners. Sorry for saying so, but chances some careless person owned it before you are quite high. Remember 30 years ago, tube testers were regarded old junk, and came in hands of people that saved it from scrap. As they cost almost nothing in those days, somebody would often randomly operate all knobs, trying to see what interesting things will happen. You wouldn't want to know about it.

So you must always check the tester's inside for repairs, and why that was done, and if it was done professional, and did this solve the problem at the root. (normally not). So I repeat my approach: If you think your tester was treated and serviced only by professionals, that is very good, but if there is no proof, any "expectations" are of little help. You as well expect the opposite. For instance if you see a solder joint, where a resistor was cut off with one lead, and the cut was re soldered, this means somebody suspected this resistor to be bad, but it was still good. You need to realize he was searching for something, and obviously it was not the resistor. So the question is, did he find it, and can you find where that is? If not, he may not have found the problem. Bad soldering is a bad sign, and components being replaced by other components from somebodies junk box, is red flag to me. Same as an opened up meter which in effect is not precise. Or even repaired with other meter's parts inside. However when you see nicely replaced components in there, new and high quality components, with well done soldering, or a sticker on the meter, saying what was done, you can be optimistic a specialist was at work.

Check those funny meter nuts. You CAN open them with a normal screw driver, but that will leave a lot of scratching. Are these scratched? This means somebody without appropriate tools has opened it. Well, good luck with that meter, but I would not be interested in it.

Protection relay. The only safety precaution for the meter, is the mains protection relay, which protects the mains transformer at appr 150mA plate current, but this will not protect the grid circuit or the heater windings, and also not protect against meter overload. It only protects the meter assuming it was set for 100mA, and you have a 500mA tube put in. It will not protect the meter if you set it for 2.5mA, and a 500mA tube was put in. The meter is damaged faster than the response time of the overload relay, and coil deformation is the result.


What is good about the AVO, there is not much overly sensitive electronics in there, no black magic, and no wire spaghetti as in the Hickoks. Other than most Hickoks, the AVO can be serviced without getting a head ache. Though the electronics work more complicated as one may expect, still in the end you will understand it, if you invest enough time. Compare that with a Hickok, I am sorry, but these weird circuits, and piles of spagetti wires, are just not my thing.

The Mk4 is the third generation testers of this kind, and very mature. As a result of many years of development all errors of the earlier models were solved. It can work on any mains voltage from 100 to 250V. It is probably the most wanted tube tester in Europe and Asia. The reasons is it's high functionality, combined with a rugged case, problem-free switches, almost no capacitors, and very high quality inside parts. Also a scanned copy of the service manual and tube data book is around. (Don't let them screw you on Ebay for a payed version). Calibration is possible on the user level, with only two pot meters, but you do need a calibration valve for this, like an ECC82 or 6SN7 with exactly know data, and you do need to READ all of the instructions. So it won't get better from experimenting, and expect some self teaching by trial and error. If you expected that, I rather advise you don't touch the calibration points.

The complicated factory level calibration is only needed when internal parts are replaced, which means adjustment of the dial wheels. (So the position of the dial wheels on their axis. Don't play around with that, as you won't understand these setting easily, and besides there only a need, when you replace a pot meter. The reverse conclusion is, when you see the set screws were touched, and the pot meters not exchanged, somebody without knowledge of the tester, experimented with it, for some abnormal reasons, whatever that was. This is an alarm signal, but call it a strength that you can detect this easily. The remaining (low level) calibration you can do yourself if you want and if you take your time.

The tester works straight forward, so you can just decide what you want to measure, and get started. You can measure an unknown tube simply by the data sheet. Furthermore the tester has excellent diagnosis methods for the most occurring tube problem: Cathode to Heater leakage causing hum. Personally when I want to know about this, I take the AVO Mk4, because the result is reliable, precise, and no study of the manual is needed. To verify if this test is working, just connect a 1 Meg Ohms resistor (or any other value) between the link A1 and A2 on the tester deck. Then do the "isolation test" with the knob on "A1" and you should read 1Meg Ohms on the meter scale, and really nothing else. DO THIS ON YOUR AVO, and if the result is not 1Meg, unfortunately you have a problem.

Isolation and shorts

The isolation tests is the finest of all tube testers I know. It is done at 200V AC. That will show corona leakage at electrodes which seem "ok" with a multi meter. Just connect a known resistor to two random tube pins. Like 10 Meg Ohms to Grid 1 and Anode. Then at position "G" and "A" the meter will say "10 Mega ohms". As simple as that. Even 25 Mega ohms you can test. With external leads I suppose you can use it to measure leakage of a mains transformer.

To make you realize what a great leakage tester the Mk4 is, I show you here a picture of a state of the art "MAVO" leakage tester. So not an AVO but a MAVO. Very expensive at those days. Look at the MOhm scale (the red one). It comfortable tests 0...10 Meg Ohms, but 30 Megs, you have a miniature piece on the scale, and bad resolution. Well this was state of the art, keep it in mind.

Now, compare this with the outer scale of the Mk3 or MK4. As you can see it measures nicely up to 25 Meg Ohms, and above that gives only an indication too. To me the scales look to have the same resolution and the same precision, it is just at the AVO this was not the main function of course. Yet, the leakage tester was not even beaten by that MAVO special leakage tester made in the same period.

Test any tube you want.

One of the best things about it, and so little to say: It has almost any socket, all connection options and all voltages.

Gas test.

The gas test is not very sensitive, but works good. However AVO-typical it works very straight forward, you measure directly the grid current in micro amps, at a 100uA scale. Now nobody is interested to see the difference between 10uA or 100uA grid current. You smash any tube with 10uA anyway. So you need to look here at the veeeery beginning of a needle movement.

The MK3, Mk4 METER

This meter is a specialty, and is found only on the Mk3, Mk4 and CT160. The meter is a 30uA with very low resistance of 3250 Ohms at the terminals. Those that have been searching for a replacement will know, such meters are not made any more. Then, it has a quite long needle, with considerable movement energy. So if you overload the meter, it violently hits the needle stop, and something damages inside. The pivot diamonds are suspended in little springs. If something damages it can be fixed with the right level of expertise, as long as it is no burned coil. In the handbook of the Mk4 is written, the movement is reasonably protected by an electronic precaution (diodes in anti parallel, and a capacitor). Well all I can tell you, it is very easy damage it, by operational errors. Also because all always the protection capacitor is fully dried up.

Today, you can still buy 30uA meters, but these have 4...6 k impedance and shorter needles.

Also the Mk4 meter is not dust-proof any more after all those years. The glass is kitted with some black stuff that is now brittle, and the meter is not closed well any more. A meter in original condition likely may not survive shipment, the glass comes probably loose, and all kind of particles come inside. Here is a nice experiment to realize how much magnetic dust is in your house. Take a strong magnet, put it in a plastic bag, and then hold it in the dust bag of your vacuum cleaner. This is random household dust, and you will see, the magnet it is full of little magnetic particles, looking like black dust. So once the glass is loose, consider the meter damaged by magnetic particles. Even if it will not stick now, later it may become another one of those infamous sticky meters. It is almost a standard repair, to clean and re-seal the housing of the meter. I always tape an opened meter from the inside, with some pieces double sided sticky tape. That will catch some particles, also after closure. Hmmm... Like a tube getter. Though it must be warned, even when you are skilled, if you open a meter, the risk damage is not zero.

The Mk3, Mk4 and CT160 meter are 30uA which is 3x more sensitive than Mk2 which is 100uA. The meter coil is mainly held in place by the circular springs, and not like people think, by the jewels. Yet at only 30uA, these springs are very thin.

The advantage of 30uA is, the grid leakage test gets more sensitive. Sometimes a meter can loose magnetism, after 50 years. So it can happen a Panel meter indicates 5% less. The crazy situation is, the user doesn't know. So he simply calibrates the mains voltage before beginning a test, and without being aware, he starts with the mains voltage being 5% too high. (Since the meter is 5% too low). The whole tester is very inaccurate this way. So the user will find out, the tester is not precise, and he will do a calibration, and this time he calibrates the tester to read 100mA on a 95mA meter etc. So now the mess is even larger, and the tester gets sold on Ebay as "new calibrated" and prices are high there. The new users runs into the same issue. When he messes up the factory calibrations of the dial wheels, that is not beneficial either. It is needless to say, such a tester is never a good one, and things get really worse if some person started to poke around inside, try tweaks resistor values, to get more desired readings.

It is said so in the service manual too, and it is true: If you can NOT get the tester in good condition, the panel meter may not be accurate. The whole procedure begins with a meter check. Personally I would say, the error triples. So if you have a 5% error with the panel meter ( either impedance or sensitivity) the tester will be 15% off for some readings, but not for all. Some readings will be fully correct. So if you change the wrong readings by adjusting something.... those readings that were good already before, will be now wrong too. If you see THIS effect, you may have a bad meter, or even some nut changed the factory settings. That setting belongs to the pot meter used. Of course, you also loose that setting when you remove the dials wheels. So if you feel you need to do that, with an Ohms meter, you should indicate for instance what is the middle of the pot meter. Not speaking of the arm-bending procedure inside the pot meter, to get it's linearity changed. This part of the factory calibration is not documented, because this is never needed to do. Yet hopefully the experts before you did not open the pot meter, and straightened that bended arm.

So don't touch remove the dial wheels, without very good reason, and always indicate their original position very precisely. And DON'T bend back the pot meter arm from the inside, when you see it was bended by somebody. More likely you change the factory calibration by that.

Better is to start with the meter!

The first test for a meter is, disconnect the leads inside the tester, but leave the meter inside. Then with external equipment, verify it's full scale sensitivity for 30uA. Second is measure the impedance, it must be 3250 Ohms +/- 50 Ohms. HOWEVER this is NOT EASY. You can not use a digital or analog multi meter, as these always pump 1mA into a 3k impedance, and your 30uA AVO meter is damaged. When you are good at it you can derive the impedance from the current and the voltage, but make sure you make no error with parasitic current draw of your meters used.

Here you see the inside resistor. With that one you can only adjust the impedance, not the sensitivity, as some amateurs try. It is just hanging loose inside. If magnetic dust gets in, don't try to brush it out with compressed air, that damages all piece parts, and the particles are still inside anyway. Also, compressed air from a big compressor contains magnetic particles from the rust inside the tank. Cleaning can only be done if you remove the magnet, which involves taking the whole thing apart, and from that other problems can arise. So a replacement meter, before you buy one, should be from an very trustworthy source.



Compare with other testers

Mk2 has the advantage of much higher negative grid voltage if you need it, but the Mk2 case is really a bit too light. Mk3 and Mk4 testers do not differ so much. Important is the nicer dial wheels with the Mk4. Same meter as my favorite, CT160 has. It is evident, that leakage measurements with Mk4 and CT160 are the finest as possible with tube tester. No other tester ever made, was that sensitive. You can easily measure 20Meg Ohms with precision even. There was a Polish tester called ELPO 508. This seems to me like the Mk5 Version of the AVO. This remarkable tester is described on this website as well.

What is nice about the Mk4 is, you can do almost anything with it, almost any tube. The extremely sensitive isolation test is seen with no other tube tester, and the "shorts" rotation knob is a method worth an extra patent. Unlike the Funke and Neuberger, the AVO M2, Mk3, and MK4 do this with a very simple wafer switch, almost no moving parts, and nothing gets defective. The Funke's do the same, but the rotation switch deck is sensitive to wear out, and needs cleaning. The Neuberger also do the same, but specially the older models all suffer from material detoriating.

Here is how I would rank them:

Ranking by the way it looks:

    1. Mk4
    2. Mk3
    3. Mk2
    4. CT160
    5. Mk1

Ranking by fast result

    1. Mk1
    2. CT160
    3. Mk2
    4. Mk3, Mk4

Ranking by precision:

    1. Mk1, Mk4, CT160
    2. Mk3
    3. Mk2

Notes: Mk4 wins the beauty contest for outside looks. If you find one that was under a dust cover always, the inside is as gorgeous too.


Notes: Mk2 wins the beauty contest for inside ingredients. It can do higher grid voltage which makes it possible to test under more real condition, and it has a real tube based bridge inside. CT160 is most comfortable and most precise of all, when you can live with grid of max 40V. Mk1 can only do Gm and not show plate current .  

Simplified check of the calibration mark on the panel meter.

I can't repeat this for all the mains voltages woldwide, and this is only for 230V

  1. Check if the fuse is set correctly. Measure the mains voltage for that. So the fuse here (at 230V) should be set at 220...240V.
  2. Set the tester for 6L6 tube, and set the anode for 200V. No tube inserted. .
  3. Insert ab AC voltmeter inside the socket pins for the 6L6 between anode and cathode. When performing a nornal anode current test, you should see there a voltage of 10% above settings. So here, 220V on the socket.
  4. Set the mains adjustment knob until you for the closest voltage. Take the setting which is "just below" 220V, and the cal mark should be "just below" too. Take rhe setting just above 220V, and the cal mark should be just above. If this is so, the tester is calibrated.
  5. When you have a variac you can work more precise. Set the mains knob in the middle position, and adjust the variac such that you have 220V AC from anode to cathode. Now, the calibration mark should indicate exactly.
  6. If the mark is not indicated correctly, you need to check the meter for 30uA full scale, and for good linearity. If the meter is good, and only then, the problem is the calibratrion resistor inside. This is a had selected value, which is composed with two resistors that are in series. How to replace those, is too much to explain here.

Simplified user's guide to the Mk4:

    1. Mains test: Set Circuit selector and leakage switch to the left position. The meter now will go at or close to the red ~ sign. Set meter on the ~ sign with the "set ~" knob.
    2. Do all settings by the book, for the tube you want to test, set back off knobs to zero, and Meter Switch to 100mA, and Insert the tube. Generally the rotary switches are now moved to the right, depending on what you do.
    3. Select Check (H). H stands for "hot" Tube will burn now. If not, stop here.
    4. There is also a way to test the filament of a cold tube, but I prefer to see it burning myself.
    5. Let the tube warm up 60 seconds. Turn the leakage switch step by step to the right. If any position shows leakage, stop here.
    6. Select " C/h insulation" to test cathode to heater insulation. Read on the upper scale in Meg Ohms. Below 400k is a problem, but most tubes will show no leakage. If there is leakage, select Check(C) and see if it goes away while the tube cools down. This confirms the leakage, stop here.
    7. If all above tests passed, select "test" and read the plate current. With the normalized Ug1 (by the book) check if the tube has the normalized plate current +/-30%. If more than +/-50% off, stop here.
    8. Check if the tube responds nicely to Ug1 changes
    9. DYNAMIC TEST (Transconductance or Quality). For this you have to set either Ia or Ug by the book. Transconductance must be tested at the plate current by the tube data manual.
      Quality must be tested at the Grid voltage by the tube data manual. Like this, weaker tubes will show a sharp fall off, nicely indicating those. This is indeed as AVO recommends it for Quality. However it is WRONG to measure transconductance as such. Now this is clear for experts anyway, but I just wanted to point this out here, because AVO forgot to mention it.
    10. TRANSCONDUCTANCE TEST: Set tube to normalized plate current, by adjusting Ug.
    11. QUALITY TEST: Set tube to normalized Grid Voltage.
    12. Wait now until the tube is stabile, so plate current is not rising any more. This can take a few minutes.
    13. What comes now os called BACK OFF Procedure.
      1. Turn the back off Pot meters (Coarse and Fine) until you have just above zero on the meter.
      2. Turn the "meter switch" step by step to the right, re-adjust the right back off again. Final adjustment is at 2.5mA. Make sure the meter is stabile now and at zero. (tube is fully warm).
        If the meter needle moves up or do
        wn slowly, the pot meter arms need to be cleaned and re-adjusted from the inside. These are the back off pots, and the Grid pot. However the grid pot has an internal calibration setting which is factory done, and "once only". So this is for specialists only. If done wrong you never get your AVO calibrated very nice.
      3. Now select "mA/V" and the meter must be in the green, and basically you're done, the tube is good. Ideally the meter is at "1" meaning 100% of the value selected on the turn wheel. At -5 it is 50% lower than on the turn wheel, etc for all other readings.
    14. You can also measure the actual transconductance precisely with the mA/V turn wheel. Just turn the wheel until the meter is at "1" and read from the wheel the actual value of the transconductance. This must have the same result as the previous method, only the turn wheel. method is more accurate.
    15. As a final control, select mA/V again, and check if the back off is still at zero. If not, you must set it to zero and start again. To prevent having to start again, do the mA/V test quickly. If you are routined you will learn to appreciate this method.
    16. Rotate the test knob to GAS, and the reading is for grid current, with the meter scale being 100uA. Do not confuse gassy tubes with leaky tubes. Leakage can be found with the leakage test. If the tube is NOT leaky, the gas test is becomes meaningful. The gas test is not hyper sensitive, but good enough.

Overall conclusion:

Perfect. Mk4 is a beautiful tester. With it's thumb wheels and universal test voltages, you can test many thousands of tube types. Accuracy can be amazingly good, when the tester is in fine condition, and serviced by an expert. It must be warned against Mk4 testers which have been messed up by so called experts, mostly Ebay experts, and internal problems were "calibrated away" instead of fixing them. Like the grid pot and transconductance pots have been set to abnormal settings just compensating some defects that they could not fix. Or a weak meter magnet was simply calibrated away by setting the mains calibration higher. (Unknowing often). Many are like this. Such testers typically give good results for some tubes, but not for ALL tubes. If the Mk4 is in good condition, backing off can be done fast and comfortable, and results are accurate for any tube you take.



Next comes the download for the manuals. Here is some quick check hints for a Mk4 tester. This is for technicians only. More about it here....


Operating Manual
Service Manual
Schematic (Very HQ scan by me. New on the internet)
Another schematic (It has errors)
Another Schematic (Thanks, Joris Weijters)
Tube Data Manual - Version 23 - Last manual printed.
Set printer menu to best enlargement that works nice.
Probably 120%. So the tube data book prints full page, and you can read it nicely