ez760 transistors and troubleshooting diary

Hello, friends,

I'm getting back to trying to get a stereo ez760 up and running. One of the problems I think I have is that when I ordered parts, I tried finding the obsolete transistors that are listed on the BOM (BC182L, etc). This was before I knew about finding modern equivalents and the existence of bad, cheap, counterfeits. I would like to know if anyone who has successfully built one of these if they would be as so kind to share which transistors you used if you did indeed swap out for modern equivalents.

In the meantime, I'm going to pick up a solder sucker as I am going have to pull all the cards that I soldered directly out of arrogance and foolhardy trust in my abilities and socket everything. The first task is to figure out why I am getting crazy high VDC at the meter signal point when it is sitting idle.

This build and it's calibration process seems to be daunting to a larger number of DIY enthusiasts, which is a shame because I think this is a great piece of gear. My hope is to be able to figure this out, then build another single channel to be of assistance to everyone here who is considering this undertaking.

Thanks!

Paul

My first unexpected item is that the sockets I have don't seem to fit the right angle header pins  that are mounted on the cards. I measured the pins to be right at 0.025"(0.64mm) which is right at the top of the tolerance for the Mill-Max and Ares that are listed on Mouser. The pins are also square which does not help. The TE connectivity versions won't show you a data sheet unless you sign up for an account on their website and the Harwin's are smaller. So If I am going to socket all the cards, which I am going to have to do, I have to find 90 degree header pins that are smaller.

Thanks!

Paul

All the cards are pulled and I've begun to reassemble whilst injecting signal along the way. With tone connected to the input header on the board, I can measure signal leaving the input card and the verify the input level control is working. Right now I am stuck as it gets to the input of the compressor card. With that card removed, I have the same RMS voltage leaving the input card to both sides of R198 on the main board, the input solder pad for the compressor card, and to the respective points on the compressor control header. When I connect the main board to the control PCB, the signal from R198 to the compressor card input solder pad drops out completely, but the opposite side R198 retains signal. Also, I can switch the limiter toggle to the "In" position and see it show up at the limiter card input solder pad on the main board. Keep in mind the only card installed on the main board at this point is the input card.

The only thing I can imagine is causing this signal to drop out is that there is a short to ground somewhere which I haven't found yet. I don't see a schematic for the control panel PCB, so that makes it kind of difficult to follow along and trace signal and see if there is something else I am not taking into account. The signal to the input pad of the compressor card is not being returned from the the control panel, so if it's not shorting out when that connection to that header is made, I don't know what it could be.

Thanks!

Paul

Output is installed and is working correctly, however I am a bit confused the with the reading from the output transformer. The (+) side of the output connector is approximately 1VAC higher than the (-) side. I unsoldered the transformer leads and connected the primaries and secondaries in series as it is in the schematic, with the brown lead connected to ground, which yielded the same results. I grabbed an unused transformer of the same variety and the same thing. Then I connected the + and - from my tone generator the primaries and measured the same voltages on the secondary leads in relation to the center tap. I had always thought that the output transformer would be symmetrical in relation of voltage to ground, but the center tap on this design does not have a ground reference per se. I blame not a lot of sleep and staring at this schematic too long hoping to see the error of my ways.

After all that unnecessary rambling, the output card is working correctly.

Thanks!

Paul

Now we install the compressor card. This is where I am still at a loss...

With the circuit engaged, the input and output controls are working fine. The limiter in switch is sending signal to the limiter card input solder pad. The limiter card and expansion card are not installed. The difference of the Meter Signal connections is about 1VDC (8V on the m- and 9V on the m+ side, don't know if either of those are correct) which does not change with the meter zero trimmer. I also do not see much of a difference when the compressor controls are adjusted. I do have all of the trimmers about halfway as I saw some other person here suggested as a helpful starting point to ensure audio was passing through. But it is passing signal whereas before it was not, which is a tick in the W column I think.

The next thing is to start chasing down voltages on the compressor card. I'm hoping to find a bad transistor and that be it. Sometimes it's that easy. Sometimes it is not...

Thanks!

Paul

Pulled the BC212B from the Q44 and put in a BC560C that I had to take from an op amp (these are on order) and now the Meter(-) voltage is changing with the trimmer showing 9.5VDC all the way clockwise. The voltage for the Meter(+) is still higher, but more like 0.4V in stead of 1V as previously stated. I also noticed that switching the attack time changed the Meter(+) voltage by 1.2V between the middle position and the outer. I may have a received a bad batch of BC212B (or even fakes!), which I won't be able to do anything about until FedEx shows up with more parts.

In the meantime, I'll just match some 550Cs for some op amps that need to be built. It's raining outside and the tornado sirens keep sounding off, so this is about as good as anything else to do on a Saturday.

Thanks!

Paul

Potato Cakes said:

The only thing I can imagine is causing this signal to drop out is that there is a short to ground somewhere which I haven't found yet. I don't see a schematic for the control panel PCB, so that makes it kind of difficult to follow along and trace signal and see if there is something else I am not taking into account. The signal to the input pad of the compressor card is not being returned from the the control panel, so if it's not shorting out when that connection to that header is made, I don't know what it could be.

Click to expand...

The schematic for the control switches is the first one in the Colour Book. I just can't read. When the Threshold switch is turned all the way counter clockwise the short goes away and there is signal present when the compressor card removed. I need a closer look at the newly discovered schematic and see if I'm smart enough to see what's going on with that guy. But at least I know there is a way to check the input and output cards without installing the other guys.

Thanks!

Paul

Time for some more.

All the BC212B's were replaced on the Compressor card and now the Meter Zero trimmer will actually zero the meter. I tried a couple of meters out of curiosity and found that it worked for a true VU, a cheaper DC type with VU scale, and a 1mA Sifam.

Moving on to the compressor section... (limiter and expander cards still not installed). I cannot seem to get any gain reduction to happen with the controls. I do not believe that limiter card needs to be installed for this to work, but I will double check the schematic. As before, all the trimmers are approximately at their halfway point, so the compressor should be doing something even if it is not calibrated.

In the meantime I will be replacing the rest of the BC212B's.

Thanks!

Paul

I started the calibration process, and with the system out, the output was reading 3.0V RMS instead of the 0.775 I was expecting per the instructions. The output card has 212B on it as well (I didn't change them because I thought it was working), so I pulled those and now I have about -6dB from input to output with system bypassed. I also swapped out the 182B's on the input card with 550C's with no change. I put one of the 212B's back in the put card in the middle position and the increased gain input was back. Swapped back to a 560C for that position and back to -6dB. I also noticed that with the System In and with the 560Cs in place on the output card, the input and output controls don't really seem to do much until the last quarter turn. Before that all the signal seems to drop most of the way out. With both controls wide open it still only gets me back to -6dB from the signal at the input. I've checked the transformer wiring but I will have to give it another look.

The instructions do indeed say to wait 30 minutes with it powered on before checking audio, so I'll do that but I don't think there will be any change. I will swap out the 212B's I have on the other board that is built up and see if I can progress any further. I am one step further in the calibration process than before, but there are still 108 steps left.

On a completely random side note, I purchased some NTE Anti-static, non-cleaning solder wick today which I was a first for me. Using this to remove solder joints emits an aroma akin to Barbasol shaving cream, which is very pleasant and fresh smelling. It causes an temporary amnesia to why one has to use this product in the first place, making removing transistors on a very crammed board with tiny solder pads less frustrating.

Thanks!

Paul

Hi Paul

I have moved over to using BC550s and 560s with no change in sound at all.

Thanks for the tip on the solder wick.

Peter

It seems that I will doing that as well, Peter. I thought the BC182B were working, but the Q45 and Q46 turned out to be bad on one of the boards, keeping voltage from getting to their respective points on the compressor card. So now I'm going to just swap all the transistors out with modern equivalents. I will be using the following:

BC182B - BC550C
BC212B - BC560C
BC184C - MPSA18 (reverse orientation)
BC184L - 2N3417

I bought MPSA92s for the BC212L, but I noticed that the 212L is ECB and the 92s are EBC. I would have to rotate the 92s 90 degrees to make it work, but there would have to be special care taken not short out the legs. I see Mouser has BC636 which would work reversed from the 212L, but I'll see if I can make these MPSA92s work. Mostly because I don't want to make another order. That will be all for today.

Thanks!

Paul

Swapped out all the transistors and the FET on the Input card and now I seem to get the +0.2dBu that the calibration process says I should see the system is bypassed. When I switch to in, the signals seems to drop completely out. However, when I crank the gain all the way I get a little more gain than before, about 1.2V RMS, which before I was getting just under unity. I found some voltages to check off of the 760 schematic that Q2 Audio has on their site, and it says that I should be getting about 5VDC between the 2n3819 on the compressor card and the CV trimmer, but I only seem to be getting about 2V despite the proper 24v going into it. Even though this FET is a modern made part, I swapped it with a new one and still the same thing. I still do not have the Limiter or Expander card installed, but I don't where that would affect the the voltage of the 2n3819.

The voltages I am getting for TP2, 3, and 4 are 24V, 1.9 (affected by CV trimmer), and 0.05V. I do not know if these are correct.

I'll finish changing out the rest of the transistors on the other cards and install them to see if anything changes. If not, I may be stuck.


Thanks!

Paul

Running some signal traces, I found there was continuity between the Signal Input on and the Audio Ground on the compressor card. There was a tiny bridge on the back of the card between the first two solder pads for the header. Amazing. That solved the gain issue. I powered it up and ran signal and noticed that the level slowly increased, which is probably why one has to wait 30 minutes before actually calibrating with signal. I also discovered that I wired my meter backwards to the Molex connector, so swapped that and now the meter is tracking with changes I am making to the compressor settings. However the output is still not decreasing in accordance to the gain reduction being displayed on the meter. Plus, on ratios 10:1 and 20:1 the meter is pegged to maximum reduction, regardless of the threshold setting. So there's a new fun one. That's where I will have to leave off for today.

I feel confident that if I were to build this now with the experience and knowledge (albeit still very little) I have obtained up to this point, I would have far less trouble getting it up and going. I did order another board so maybe in the next couple of months I will get to put this newly founded confidence to the test.

And back at it...

I retested the board that I had previously solved the gain issue to make sure it was still as I had left it and the problem has somehow returned! Very frustrating, indeed. I grabbed the other board to double check it and it had no gain issue. The meter also tracked properly with the varying compressor settings, however no actual compression of the signal occurs. This one still needs the BC182s replaced, so I'll give that a go and see if that gets me closer. Still a little bummed that the problem I thought was resolved has come back.

The new board arrived so hopefully that will be of some help testing continuity between points.


Thanks!

Paul

More test results.

The FET on the compressor card is changing voltage to the mix point based on the adjusting settings to the compressor controls, so I believe the card is working properly now. What I didn't know is that the the mix point passes through the the Expander card via the control card. So now I need to make sure all the transistors on the expander card are changed out and double check all the soldering before installing it. The Limiter card is still out of the circuit at this point for the board I am currently testing and it looks like that according to the schematic I do not need it installed to see compression working.

Standby for partial excitement!

Thanks!

Paul

It's alive!

It seems that I was very wrong about needing the expander card. Apparently, there was no compression because of the Inverter trimmer not being set correctly, even though I thought I had changed all the trimmers to their halfway point so that everybody was doing something that I could measure. There still may be some fun things to find when I start calibrating, but the VAC output is changing with adjustments to the  compression settings.

The other board has an new shorting issue with audio and ground on the control PCB, which should be easy to find.

Time to put everything back in the case and calibrate!

Thanks!

Paul

Managed to get the second board up and running as well. I found that the connection from the Inverter output wasn't getting to the FET control on the Input card. It's a long copper trace from one side to the other and the connection from the compressor card to the through hole jump that connects the top and bottom traces for this particular run, but it wasn't getting to the input card, so a long jumper was in order.

Calibration will have to wait as it is well past my bedtime. I've read that matching the Expansion controls for stereo use is a little tricky, so I'm looking forward to that.

But as is, but boards are compressing, limiting, and expanding. I'm relieved that this project will see completion pretty soon and not be a constant reminder of my failures...

Thanks!

Paul

Started the calibration process and everything seems to go fine until the expander part. The trimmers on the card didn't seem to do anything. I swapped Control boards and the same results, so I can rule that out. I had some more of the BC212L that I changed out for the MPSA92's and I managed to get the process started up the the FET Law adjustment. For that the meter stays bottomed out at -20dB. Also, the LED goes out around -3dB of reduction and not near the 0dB in the initial Threshold adjustment. I checked the schematic and I noticed that there are different transistors between the AML schematic and the one that Q2 Audio provides with their ADR760 product. The connection points and the values of the surrounding components are the same, but the Q2 schematic has BC182B's and 212B's where the AML schematic has 184L's and 212L's, which are not directly interchangeable. I'll have to do some reading on the original ez760 thread (which I think is pretty much dead at this point) to see if this is the case. If I have time I may just use the transistor values on the Q2 schematic for the expander card and see if that makes Bob my uncle, who is already my father.

One fun tidbit is that the Input Threshold trimmer for the compressor card is not mentioned at all in the calibration process, which needs to be adjusted to something otherwise everything else downstream during the compressor and limiter adjustment procedure will be adversely affected. I landed on having 8.33k ohms between the wiper and the top facing leg of the trimmer as being a good starting point as it allows me to achieve victory for each step up to the expander section.

So very, very, very freakin' close!


Thanks!

Paul

Just realized at 4:30am that the Q2 PCB probably accounts for the pinout of the 182L versus 182B (which are still current and probably why they are being used in a production piece). Same goes for the 212L/212B, so doing a direct swap on Colin's board is probably ill advised.

I saw someone else having difficulty on the FET Law adjustment part, as well so maybe I just need some sleep and come back to this tomorrow...

Thanks!

Paul

I went through the calibration again with the Expander section and I still can't seem to get the FET Law part to work. On the other board I cannot start the expansion calibration process at all, so that's not frustrating. If anyone who stumbles upon this thread has some tips for the last part of calibration I would gladly take them.

Now I need to try to get a couple of hours of sleep before I have to work for money.

Thanks!

Paul