Follow us on Twitter!

© COPYRIGHT NOTICE All Rights Reserved

(Last updated: 12-May-2018 11:58 )

Here are some nice instruments listed for tube DIY



  1. Next

Multi Meters

In some applications an analog multi meter makes no sense, but there are also application where they are comfortable, or even the only right way to go. Such as calibrating the famous AVO tube testers. AVO write in the manual, all voltages were measured with an AVO8, and you should not use another. The AVO 8 gives a small load to the circuit, which is included in the test result, and it indicates the mean value of some pretty wild AC wave shapes inside the tube testers. So you can not measure that with an RMS digital tester, and I would never be certain if an electronic Average value is the same as the AVO8 result. Not saying the AVO8 is perfect, but you do need the same behavior, or you are calibrating your AVO tube tester the wrong way. So that's a typical example where you can hardly replace an analog tester by a digital one.

Another application is, when it comes to low current DC measurements. You probably will soon find, digital meters perform only good on paper, but not in reality. I have seen a great variety of multi meters, when I was working at Agilent, and when you want to do precise current measurements, in the range of a few uA only, you can forget about most of the digital meters. Their specs are fine, but there is no use is reading "3.61uA" for instance, when the last digit (so the "1" in this example) is constantly flickering, and you can not even see what the meter is trying to say. Whereas the "6" is changing all of the time from 5 to 6 and 7. So it's a 6, but what a silly way to measure this. Then, I take my analog UNI-10 and I can exactly read on a full scale of 10uA the value of 3.6uA and estimate the next digit even. Next is, when you want to see a tiny difference. Those constantly flickering digital numbers are useless in such a case, and very fine needle movement is much more what you need.

Electronic multi meters with an analog panel meter are the best for searching for hum problems in an amplifier. I have an Unigor 6, and it has a full scale of 1mV AC. It's mirror panel meter is big size, and it has a tout band suspension, for frictionless movement. This means you have 100 scale divisions which you can nicely see, and which are reliable and correct. In between you can estimate 1/2 division easily. This means you have 10μV per division. And you can easily estimate 1/2 or 1/3 of that reliable. So a resolution of 3μV, and it's a stabile, non flickering value unlike with a digital meter. I think a digital meter which can display a stabile (so non flickering) resolution of 3uV doesn't even exist. If it does, you can let me know.

Definitely not good with analog multi meters is the absolute precision. You often have to work with them as is, which is like 2% or so, when they were new, and aging or panel meter damage, adds another 1..2% to it. They have often no calibration possibility, and that becomes very tedious then. However some meters can be amazingly good still, just as they are, and you do good to search for those who have a tout band suspension of the moving coil. Such have no aging of the movement, no friction, no stick-slip effect, and coil deformation is very rare with those too. Those meters from the 1960--1970's using those good old East German tout band meters are the best ever made. The best I have in my collection is the UNIGOR 6e. I keep it as a treasure, to do sometimes very delicate measurements.

EAW East Germany

EAW East Germany. Not so interesting looking, but.. this is a tout band meter with mirror scale, and that is usually a very good meter. This one I used for calibrating AVO equipment, because it has the same impedance as an AVO 8 multimeter. I have finally found a good working AVO8, so I sold this EAW meter. It has a one calibration pot meter inside for AC and one for DC in general. This calibration option, 99 out of 100 meters do not have. So it was designed with precision in mind. It was easy to drill a hole in the deck, and then I could can calibrate this from the outside. This gives the option to calibrate for instance a particular DC range, let's say the 30V range, and then use this for a calibration measurement of an AVO tube tester, where I need the load and averaging effect the analog way. Since the meter looks pretty ugly, nobody even looks at it in Ebay, and I bought it for 7 Euro.

What is nice, it has separate banana connectors for voltage and current. (So the four you see at the top) and you can hook it up to a circuit and change from voltage to current measurement without having to plug cables. When you go to "voltage" it shorts the current input bananas. When you go to "current" it opens the voltage bananas, for "no load". Really nice done. Like this, it becomes hard for a fool, to swap current and voltage by mistake. I belong to this category too ;) Also there is a security button you need to unlock before changing to another sensitivity, which works simple and good. So chances on buying a damaged one is low.   Make sure the case is has no cracks, and the bottom is not scratched as it has printed quick use guide on it. This is just a very nice tester for a low price.

UNI10 by MTM East Germany

Picture from the Back
Picture from the Side
Manual pdf


UNI10 by MTM Messtechnik Mellenbach Thueringen VFB / DDR. Build in 1978.

You have to digest the specifications of this meter, before you look at anything else. MTM is a producer of tout band panel meters, and they must have use this situation to build the most sensitive passive multi meter build ever. It looks not attractive, and we learned on Ebay, failing sexy looks makes the price low. You can find an UNI 10 anywhere from 20 to 100 Euro, depending on it's condition. You need to know it was the workhorse in the technical departments of the "People's army" of East Germany. So sometimes a seller has more than one for sale. The ones that are directly from old depots, are all kicked around, and it must have been a habit by "People's army" to store instruments dirty and humid. So don't buy any of those with a dirt on it. Also since there are many UNI7 etc around, so other than UNI10, people don't look very well what they are buying. Be pickery, and buy a nice looking UNI10, and you get specs you have never seen before with a passive multi meter. Here are just a few of the specs:

Resistance: 100.000 Ohms/Volt. So at 100 Volts you have a 10 Mega Ohms load.

DC Volt: from 100mV (!) to 1000V at 1.5% accuracy. You can't outdo a digital meter when measuring normal voltages of course, but look at that 100mV scale. It has a resolution of 1mV. Not all digital meters can do that.

DC current: 10uA full scale. Many digital multi meters have fake specifications. They do show a digit for 0.1uA steps. Only when you try to measure for instance 50.6uA, you will see the "6" is blinking between all kind of numbers, and you may not even be able to guess it's a "6". With the UNI-10 this could be read directly from the 50uA scale. So 50 plus 3 divisions. The only affordable digital meter which does not blink the last digit is the Hewlett Packard 3465A. (For this please do not refer to manuals. Stability of the last digit is specified nowhere. So take a real meter, and try to measure 50.6uA and look if the last digit is stabile. And if not, you can not seriously use such a digit. For sales purposes, like in a food shop, it is not even legal to display an instable digit.

Capacitors: from 1000pF to 2uF.

Excellent Protection: Active and passive. Activates at 15x overload or at wrong polarity. Also it features a vacuum spark gap to protect against very high voltage, with virtually no delay. The passive relay protection works as an emergency solution, and it works with a reset button. So when you have pressed it, there is energy stored in a spring, which is used for reasonable fast operation of the relay. However you can't operate that at 15x overload of 10uA. So for active protection which works at low current, or works really fast at high voltage, you need to have a battery installed. The battery is also needed for Ohms and capacitance measurement. The transistor protection circuit is always "on" but passive current drain is less than the natural loss of the battery itself. This circuit allows immediate switch off in case of overload. The connections to a capacitor under test are then cut also.

Many of the other specifications are just average, but I use my UNI10 often for tube grid current tests. Which is nasty because you try to measure a few uA at a faulty tube, and when the tube fails during that, it can blow up the meter if there is no good protection.

Disadvantage: There is no direct reading possible, you always need to divide or multiply the numbers on the scale by 2. So it has "50" as a full scale, but that would either mean 25V or 100V depending on the range you select. Every time I use it, I wonder why they did that.


AVO8 by AVO. Several Millions were build from 1923 to 2008.

I think this meter is over hyped, but it has specific use to me. First, they are needed to calibrate an AVO tube tester, and have exactly the right load, and right indication of the average value, with strange wave shapes. Apart from that, this tester is extremely useful in tube land, as it has a 2500 Volts range. In case you are interested, is a bit strange meter to use, and there are many, very many AVO8 around with a deformed moving coil. This can not be repaired, and it gives unlinearity. The different versions are from Mk1 to Mk7, and definitely beginning with Mk5 they look nice. However the Mk5 is very problematic to repair. They used a plastic foil assembly instead of normal wires, and it is difficult to take that off, as it melts easily. . However you must take it off for most repairs. Next problem is, the Mk5 has those nice looking red collars around the reset button and the "reverse" button. However that is bad quality rubber, and it cracks. Some users peel it off then, and it looks "normal" again. The problem is, the reset button is physically part of the meter system. So when the red rubber is cracked, or peeled off to sell it on Ebay, you have now an air path directly from the front deck into the meter, and magnetic dirt will come in for sure. Also with Mk5 and higher, the resistors inside are laser trimmed, but the assembly is not painted or protected against detoriation. So precision of the laser trimmed parts is either already a problem, or will become a problem later. So the highest version AVO8 I would recommend is the Mk4. Having said that, I would never take Mk3 or lower, because they are just too old.

Metrix MX573

New Generation Product

Metrix MX573. Tout band meter with very precise mirror scale

This is just a nice multimeter. Metrix still exists as the joint company with Chauvin Arnoux. Until 2012 some stock left over was still normally for sale via the RS electronics website for 275 Euro. I don't know the production data of those, but I guess they must have been 15 years old or more. They were unused, and sold as such. I tried calling with Metrix France to get the calibration instruction, but they were very confused by the question, and surprised they had such a product for sale. Some bla bla was send by email about by the support department, but not what I needed. Then I found the calibration instructions just it in the internet somewhere.

I like using it for it's high level of comfort. For the rest it doesn't excel with brilliant electrical specifications, but specs are not bad also. Very good is remarkable low battery usage, and a full electronic protection of every range. So theoretically you can set it for 200 Ohms, and plug it in the mains by mistake.

Indeed these need calibration. I have three of them, so you can see i like to work with it. They do seem to have some internal noise, meaning at high sensitivity AC, it doesn't go fully to zero, but it's just a tiny little bit, and no problem.

I have the schematic and calibration instruction in French. What is interesting, the digital meter is a separate unit with an analog input, just in parallel to the analog meter. The LCD display is a standard device, and the digital part it a single chip IC, which original chip you can still buy on Ebay as surplus. Also complete units could be tried, as the original unit is 200mV full scale, and that is still normal today. So try to get one digit more, with a newer version digital read out. MX573 was made as an industry product, not a toy for the DIY market. in Germany sells the second generation of those, under the new brand name Chauvin Arnoux now, for the nice price of 541 Euro. (Year 2018) The specs are comparable, but it misses the nice, high resolution mirror scale of the MX573. The new product CA5011 has a digit digital read which can display maximum 3.999 whereas the old product can display maximum 1.999. An improvement, but not so large. So at 4.321 it can not display that, and would jump to 04.32. This is the same weakness as with the MX573, but at least the MX573 has still the tout band panel meter with mirror scale, and 100 divisions. Whereas the CA5011 has a cheap panel meter, actually of no real use to read something with it, resolution is too low. Well we need to be realistic, a mirror scale, tout band, panel meter, with 1% error (so 100 divisions) such as MX573 has, costs at least 100 Euro directly off the factory in quantities of a few 100's. If such a factory exists at all. Probably not! So I understand the situation, but then I rather keep my good old MX573.

The Ohms measurement is done nicely, because it works opposite direction as with low cost analog meters. So at higher resistance the needle moves to the right. Meaning at open connections it moves all to the right, at "endless". Which is more logical as soon as you get used to it, and then you don't want anything else.

Unigor 6e


Unigor 6e. Tout band meter with very precise mirror scale

This is the best analog multi meter I know. It is not impossible to find, but a bit rare still. People do not really know, they just see a clunky meter, and are not interested. Also there are many other Unigors for sale which look just like the 6e, and people see too many of them. The 6e however features a 1mV AC scale, with a beautiful 100 divisions mirror scale and a friction-free tout band meter.

Battery consumption is very low. This is best described by the way I bought it in 2012. It came from a military surplus store, and inside were military Batteries of the type NBA3042. 4Ah 1.5V Type C, "Baby" cells. They were still full. Is that a surprise? Well read the next item below here.

The Unigor 6e is not just precise. It is specified for 1%, and that's what it really is. All ranges, anything.The electronic protection saved it a few times from my own mistakes, and probably also from previous users mistakes.

The unigor 6e features a FET chopper amplifier, which chops the DC signal into an AC signal, and after amplification it gets rectified, so you have a DC signal again. That way it can be amplified without internal DC offset or drift problems of the internal amplifiers. Also amplifiers can be made to block anything below the chopper frequency, so mains hum doesn't affect it either. Amazingly the internal parts are not even that much, and the PCB material is very old fashioned. Yet it's functioning is really good.

Interesting feature of electronic multi meters is normally, the scales for AC and DC are the same, as they use electronic rectification (not rectification with a diode, which looses the forward voltage). So it has a linear scale for AC too.


1990 Batteries. Still good!

Sharp Battery,  still not leaking
1988 Battery. Still good

Sharp Battery,  still not leaking
Oxford 1840 battery

About old batteries...

We are getting so used to it, that unused batteries are leaking after one or two years, that we think it is normal. Does it have to be like that? I think not! Here are two curious examples of batteries im my possesion, that refused to leak empty. Even though they had seem a lot a use. The use was at very low current consumption of course, but this makes me think, using very tiny little current impulses from a battery, does "the" magic thing to prevent leakage. Also in the "Oxford Electric Bell" battery from 1840 I see the same effect. So it produces a very tiny current spike, very short only. And that goes on ever since 1840. This is not a joke. Google for Oxford Electric Bel. The following two examples, the Unigor 6e batteries were also in use ever since, but the Unigor 6e draws almost no current when it is switched on. Same for the Sony Remote control batteries. Current draw is just a few mA when you push a button, but the buttons are pushed in total only 10 seconds or so, per day. So almost nothing on yearly base, and it is neglectable in realation to the capacity of the full battery.

Found in my Unigor 6e. These batteries were used since 1990 and still good.

When I bought this Unigor 6e in 2012, it had 4 batteries inside, and it was working. After checking them, one was 0.5V the other three 1.45V. So it was running on 5 V which is above minimum 4V specs. In the bottom was date code November 1990. Isn't that crazy? I took them out, and saved them as curiosity. One of the three is getting weak now, but two give still 1.45V. Just remember what you have been doing in 1990. Well I have some still working batteries from that year.

Found in our old TV Remote control. Probably from 1988.

These were in a Sony TV remote control. Bought my wife when she was a student. It must have been early 1990's she said. We used this TV, until in 2016 the analog TV antenna signal was switched off here. So we had to throw the TV away, but it was still good. I took out the batteries of the remote control, and look what came out: The original batteries, never replaced ever. They are still good, while I write this in 2018. So almost 30 years old batteries. On the batteries bottem is a datecode 88-10.

Tektronix Current probes


The Chinese sweep this market empty- They use them massively for designing switched power supplies, which is their specialty. So supply is low. Sometimes you see them offered from China on Ebay, so that's a device they didn't want to have any more.

This is no insider report, for those who know all about it, and are looking for more. This is rather if you ever thought of using a current probe for DIY projects, and don't know where to begin. At least I had this problem myself. It seemed quickly clear: Tektronix makes and made the best current probes ever. They are smaller and higher bandwidth than any others, and it seems to me the vintage probes are much sought after, though they are 50 years old or more.

A probe should look not like an old tool, but like a carefully treated instrument. The plastics of the head can damage, if forced with violence over too thick cables, and that can chip off that little sleeve inside which is glued on the Hall element. So beware for cracks or chipped off pieces. The typical death cause of a probe is a cracked hall element, which can't be fixed. So there is a reasonable supply of cases upper part, lower part, cabling, connectors. etc etc.

Why use a current probe?

There are direct ways to get the current waveshape of a device on the scope screen, like use a series resistor. That becomes complicated if the device is connected to a high voltage, and you can't ground your scope there. So then you need to make a differential measurement, which myself I find very easy to do, but for some reason most people regard it something difficult. Consider a scope with differential inputs otherwise, such as the good old HP130C (some say it's the last good scope, HP made). Or several of the Tektronix, but these are all a bit of a Dinosaur. Yet, there is no way around opening up the circuit, and put a series resistor in there. This of course is tedious for fault finding, or doing random tests on all parts. So what is the reality? It doesn't get done. So here is where a current probe comes in. There are so many comfortable things you can do with it. Like a bundle of wires coming out of a transformer, but which one is which... That sometimes would be instantaneously clear when you know the current and the waveshape. Or have you ever thought about the AC ripple current through a power supply capacitor? When offering a transformer Ra of 4600 Ohms, where the circuit diagram says 4400 ohms, people react as if this something very wrong. Then, they put a 200uF cap on a position where 30 uF is maximum allowed, and do not even think about what that does to the transformer saturation, to the choke or the poor tube rectifier. When using a simulation program like PSU designer, you get an idea of why that is very important indeed, and why the current or peak current can be out of specs so quickly. What a simulation program can do, but very inaccurate, is check the start up behavior of the power supply. This and many other situations is just asking for a decent current probe. Have you ever heard that deep noise, coming out the amplifier's mains transformer when you switch it on? The excessive current draw sometimes even visible via the room electric lights. Do you remember it? . With a current probe you can an least see the primary current draw. That will shock you, to see so many Ampere go in there for a peak of 0.2 seconds. At the secondary side, you can see which windings are the users of the excessive current, and follow the path from there to the end user. So which part is absorbing this massive energy for 0.2 seconds? Perhaps the choke saturates, because you made the first capacitor too big, and you are slowly killing the rectifier tube with it. No trying to be dramatic here, but so many causes and mistakes, can become visible now. If you build an extension socket, from isolated pieces of 2mm copper wire, you can measure DC plate current of a UX4 based tube, without opening the amplifier. Quite useful actually. There are so many applications for it.


What probe to choose?

The first choice is, there are DC probes and AC probes.

1) AC probes. The types to choose from are several, P6016, P6019, P6020, P6021 and, P6022.

P6016 is the first type, and like all probes, the passive response is too poor for Audio applications, so the 3dB low point is far too high, and sensitivity is too low to display signals of 1mA. Together with it's dedicated amplifier R131 the sensitivity becomes 1mV/mA at a 50mA/div scope, so a 5mA/div scope like the Tektronix SC504 which I am using, results in 10x the sensitivity, or 0.1mA per division which is really very nice. However, the low 3dB point of 50Hz makes I can not measure power supplies with it. So P6016 is useless for me.

P6019 is better than P6016. With the (somewhat) universal probe amplifier R134 it acts basically like the P6016, just with higher bandwidth. It has a low 3dB point of 12 Hz, and a 1db point of 21Hz.So this covers really the audio range, and AC parts of the power supply too. The high frequency 3dB point of the P6016 was already 20mHz @ 3dB, but P6019 comes even with 40MHz @3dB.

P6020 is the companion of P6019. It is made for higher frequency, and the technical limit is set by the R134 amplifier, not the P6020 itself (which goes up to 200MHz with passive termination). The disadvantage is, the low 3dB point is higher, we are at 50Hz again. So P6020 is useless for me.

P6021 is similar to P6019, but it features an extra contact which connects to the amplifier 714A, and allows the amplifier to identify the probe. The 714A fits only in 7000 series scopes. It can however also be used on ab R134 amplifier.

P6022 is the high frequency version of the P6021, so we can't use it.

Use the R134 Amplifier with the P6019. This little device changes everything. Now, the P6019 will go down to 21 Hz fully linear, and from there to 12Hz @-3dB, and fully collapses below 7Hz. Also sensitivity is excellent, and you can see signals if 1mA with very good resolution. Amazing is also to see the transformer based device work at such high frequency. I have to admit, my equipment can't hold up with it. So it's specified up to 50mHz? Yes, it sure looks like it, when I can display a square pulse of 30nSec. It really stresses my 80mHz scope to make this visible, and I must say about the R134 amp + P6019 combination, the pulse is displayed free of distortion. When I look at it directly on the scope, I see no difference with the current probe signal. Just a few nano seconds time delay.

So yes, they're only AC, but they go deeper down in frequency than any loudspeaker, any AC signal in an audio your amplifier or power supply.


The R134 Amplifier is good thing. This is a high frequency amplifier with bandwidth up to 70mHz. It is made really nice, with an external power supply. So it is small and light weight. It features a clipping connector, which makes you can hook in up on any scope directly on the input. So to say, you add a current probe input to your scope that way. I am afraid to damage the scope like this, and I stick a cable-to-connector in there, so I can hook up a normal BNC cable on it. The output is 50 Ohms, so it will drive a cable without reflections. At the scope end, you need to use a 50Ohms termination, and then you can make indeed needle impulses visible, in the range of 30nSec, with the probe. For low frequency the 50 Ohms termination is not needed. The good part of the R134 is, there are too many available. So when it damages, you just buy another one for a low price.


A DC probe is made of an AC probe plus a Hall element for the DC and low frequency range. The first one, was the P6042, and it's still in interesting device, provided there is nothing wrong with it. The advantage today is, you use it as a separate unit on the bench, it simply has an output for your scope, and when you are done, you can put the probe inside, in a shaft. The disadvantage however is, the amplifier is complicated, and they do get defective, and many parts to adjust. The probe itself is not so much different from the later family, but not compatible. Though the transformer assembly is comparable, you always have the probe + amplifier as a combination.

The Hall elements are quite variable. Some have negative offset, some have positive offset. The greater part of the offset is compensated inside the probe with hand selected resistors. So this means you can not just replace the hall element inside a probe, by another. In that case you also need to re-adjust the balancing Inside. The final DC offset rest is adjusted to zero inside the amplifier itself.

When there is something defective with the probe it's normally the Hall Element itself, as these are the ones that get sold on Ebay. There no replacements around. Some real specialists operate the Hall Element out of a later 6302, but offset of 6302 Hall Elements is larger then P6042 Hall elements, so that is tricky. Also 6302 has two versions, of which you can only use the older version, with circular arranged pins, and not the one with a rectangular arranged pins. So consider a P6042 carefully. These seem to go for prices between 400 and 700 Euro depending on the quality. Myself I would rather pay more for a very nice one. If they studs at the back are cracked, it has been kicked around. Check the probe for cracks. If the probe screws look as if they have been taken out a few times with a non fitting screw driver, that is what it is So taken apart multiple times, but there are no serviceable parts inside, and they did not know how to use a screwdriver. Not the kind of service I am dreaming of.

Probe maintenance (That is for all probes).The magnetic core may not be touched with your fingers, and if so you must clean it with alcohol. Old cables can become stiff, until a point where it becomes very inconvenient, and you need to replace it. There are cables for sale on Ebay, good used ones, and sometimes NOS. I would not play with "any" cable, so you don't have one with the wrong capacitance or whatever. 50mHz is really a lot, and it would be a pity to compromise on that, just because of the wrong cable. The moving parts can be lubricated sparely with silicon oil. Do not use any other oil or grease, as that may crack the old plastics later. For service you can unscrew the BNC connectors, this is not a crimped connector.


The DC probes P6302 and A6302 do not differ so much from each other. The A is the later series. Inside can be a Hall Element assembly with rectangular connector or with a square connector. The square connector is the later version. I do not know how this relates to the A or P series.

These were made to be used with a universal amplifier, or with a passive termination device. Logically with an amplifier sensitivity and bandwidth is a lot better. There are a few amplifier types around. Long ago, these were of course sold in a ratio of 1:1 probes + amplifiers. However the probes are delicate, and today there simply too many amplifiers for sale. So the amplifiers are cheap and the probes are expensive. Unlike the P6042, when you have a problem with the amplifier for the later 6302 probes, you just buy another amplifier on Ebay at a low price. When the probe is good, it always works on a random, good amplifier.

The amplifier AM503 is a unit which fits into a chassis. These mainframes can be bought with 1, 2, 3, 4 or 6 slots. So when it's a "current probe only" application, you can buy the TM501 chassis. Or, you may want to plug more units into the chassis, and you need the TM502, 503, etc. I prefer the TM504 because I can still carry it, and it has no fan. The next generation chassis is TM5000 series, which has a GPIB interface, 4cm higher size, heavier, they have a fan, and they're not fully compatible with TM500 modules. I expanded the 4 slot TM500 chassis with an 2 channel 80MHz scope, and the low distortion oscillator SG502. The scope can be externally triggered via the back panel with the oscillator, and by itself this set up is a great tool, for working on an amplifier, and it features a 50MHz current probe now. It's all I need :)
Power Supplies

Hewlett Packard P6236 power supply

Today my Hewlett Packard 6236 Power supply went up in smoke.

Hewlett Packard P6236 power supplyRead why using 220V devices on "so called" 230V destroyed it.


Follow us on Twitter!

© COPYRIGHT NOTICE All Rights Reserved

(Last updated: 12-May-2018 11:58 )