DB54 - 2254 style compressor {update 22.10.19}

[MoJo]

Some pics from my units which I built in 2019. Everything works as it should except for the meter tracking (which was reported before from other users as well). I used BC184C, BC214C and BC441 for the transistors. It might help.

I used BC184C (with 2 perfectly matched ones where specified), BC214LC (could not find BC214KC anywhere), and I don't know where the BC441 went so I can't tell you what I used there. Thank you Undos, every perspective helps to rule out a potential variable.

 

[Undos]

Transistor orientation looks good now on your unit. also no other errors detected at first glance. as far as the bypass off/on position is concerned, according to the instructions under point 6, the up position (LED green) has the compression switched on. but on my device it is the other way round, the compression is switched on in the down position. i am pretty sure that the instructions are wrong. I would assume that compression is switched on when the toogle switch is pointing downwards.

 

[MoJo]

Transistor orientation looks good now on your unit. also no other errors detected at first glance. as far as the bypass off/on position is concerned, according to the instructions under point 6, the up position (LED green) has the compression switched on. but on my device it is the other way round, the compression is switched on in the down position. i am pretty sure that the instructions are wrong. I would assume that compression is switched on when the toogle switch is pointing downwards.

I may have put the LED in backwards Undos; believe me when I say one of my nicknames is "Wrong Way Joe"! Hahaha!

 

[JMan]

I discovered something odd on my board. The bi-color LED appears to be of the common-cathode variety, meaning the toggle switch should determine which anode sees current. However, on my PCB, there is no connection between the lower pin of the switch (presumably compression OFF) and the second anode of the LED. In the picture, you can see that a partial trace was clearly laid in the design software, but it doesn't end up anywhere (and no, there's no via, and I have confirmed this lack of connection with a meter). So even though it is supposed to support a bi-color LED, in practice I'll only be able to use a single color or I'll need to make that connection with a wire. I wonder if I got some early revision of the board and this error was corrected later, since it sounds like others are using the bi-color LED without issue.

 

[MoJo]

I discovered something odd on my board. The bi-color LED appears to be of the common-cathode variety, meaning the toggle switch should determine which anode sees current. However, on my PCB, there is no connection between the lower pin of the switch (presumably compression OFF) and the second anode of the LED. In the picture, you can see that a partial trace was clearly laid in the design software, but it doesn't end up anywhere (and no, there's no via, and I have confirmed this lack of connection with a meter). So even though it is supposed to support a bi-color LED, in practice I'll only be able to use a single color or I'll need to make that connection with a wire. I wonder if I got some early revision of the board and this error was corrected later, since it sounds like others are using the bi-color LED without issue.

I am able to verify that I get green (switch down) and red (switch up) colors, so I may have a newer PCB, JMan.

 

[JMan]

Who knows. I've been reading back through this entire thread and while there's some positive feedback, the project does seem to be plagued with gremlins for many people, exacerbated by a very glaring lack of any schematic (it is asked about again and again throughout the thread and the question is always completely ignored). So, I'm not surprised to find a minor error on my PCB. Somehow that feels like it fits. Part of me is thinking I might put this back on the shelf. I don't want to drop cash on the transformers just to end up with a unit that is full of bugs. We'll see.

By the way, Mojo, shortly before your first post in the thread, someone else had the same problem with the unit only passing signal in one mode and not the other. His question was never answered, but it might be worth trying to get his attention to see if he ever got it sorted out.

 

[MoJo]

Who knows. I've been reading back through this entire thread and while there's some positive feedback, the project does seem to be plagued with gremlins for many people, exacerbated by a very glaring lack of any schematic (it is asked about again and again throughout the thread and the question is always completely ignored). So, I'm not surprised to find a minor error on my PCB. Somehow that feels like it fits. Part of me is thinking I might put this back on the shelf. I don't want to drop cash on the transformers just to end up with a unit that is full of bugs. We'll see.

By the way, Mojo, shortly before your first post in the thread, someone else had the same problem with the unit only passing signal in one mode and not the other. His question was never answered, but it might be worth trying to get his attention to see if he ever got it sorted out.

I also saw the frequent requests for a schematic JMan.

I started looking for data by troubleshooting using the BA185 schematic. My limited schematic reading skills indicate that Pins U and V are feeding the 31267 style transformer, and Pins R and T are the output of BA185, feeding the 10468 style transformer. Here are my values (I am feeding a 1kHz sine wave into the XLR Input jack:

Compress switch in down position:

Pin U = 2.04VAC
Pin V = 2.02VAC
Pin R = 0.003VAC
Pin T = 0.003VAC

Compress switch in up position:

Pin U = 0.003VAC
Pin V = 0.003VAC
Pin R = 0.003VAC
Pin T = 0.003VAC

Does this look reasonable? If it does, then I will move on to the next board.

 

[MoJo]

I also saw the frequent requests for a schematic JMan.

I started looking for data by troubleshooting using the BA185 schematic. My limited schematic reading skills indicate that Pins U and V are feeding the 31267 style transformer, and Pins R and T are the output of BA185, feeding the 10468 style transformer. Here are my values (I am feeding a 1kHz sine wave into the XLR Input jack:

Compress switch in down position:

Pin U = 2.04VAC
Pin V = 2.02VAC
Pin R = 0.003VAC
Pin T = 0.003VAC

Compress switch in up position:

Pin U = 0.003VAC
Pin V = 0.003VAC
Pin R = 0.003VAC
Pin T = 0.003VAC

Does this look reasonable? If it does, then I will move on to the next board.

BTW JMan, I saw that person's posts and am trying to follow the suggested troubleshooting process, but cannot determine whether a 1kHz sine wave equals 1VAC. So I am just recording voltage values going n and out of the boards to see if you or anyone else notices a potential problem. The numbers I am getting don't really say much to me, being ignorant of the design in general.

 

[JMan]

1kHz is just the frequency of the tone - it's a common frequency for a test tone. The level you feed to the unit, rather than the frequency, will affect how you interpret subsequent VAC readings in the circuit. A good reference point is 0.775VAC = 0dBu, and 1.228VAC = +4dBu = 0VU (in many typical studio applications, although 0VU is not a true standard value but a reference level). This will not always have a consistent relationship to dBfs levels displayed within a DAW, but I seem to recall that the last time I was calibrating something a tone generator plugin set at roughly -18dBfs gave me 0dBu on my interface's analog output - I could be misremembering though. You can measure AC voltage between pins 2 & 3 on the XLR input to see what you're actually feeding the unit and adjust it accordingly. Maybe try starting with a 0dBu/0.775VAC signal.

I would say that 0.003VAC is effectively 0VAC, for all intents and purposes. So with the unit engaged, I would think having that on pins R and T might indicate a problem. That would mean that, assuming your reading of the schematic is right, no signal is reaching the 10468 transformer. (Sorry I am not double checking the pin out as I write this, I am on my phone and it's a little clunky to try to look through schematics on here).

 

[MoJo]

1kHz is just the frequency of the tone - it's a common frequency for a test tone. The level you feed to the unit, rather than the frequency, will affect how you interpret subsequent VAC readings in the circuit. A good reference point is 0.775VAC = 0dBu, and 1.228VAC = +4dBu = 0VU (in many typical studio applications, although 0VU is not a true standard value but a reference level). This will not always have a consistent relationship to dBfs levels displayed within a DAW, but I seem to recall that the last time I was calibrating something a tone generator plugin set at roughly -18dBfs gave me 0dBu on my interface's analog output - I could be misremembering though. You can measure AC voltage between pins 2 & 3 on the XLR input to see what you're actually feeding the unit and adjust it accordingly. Maybe try starting with a 0dBu/0.775VAC signal.

I would say that 0.003VAC is effectively 0VAC, for all intents and purposes. So with the unit engaged, I would think having that on pins R and T might indicate a problem. That would mean that, assuming your reading of the schematic is right, no signal is reaching the 10468 transformer. (Sorry I am not double checking the pin out as I write this, I am on my phone and it's a little clunky to try to look through schematics on here).

Thanks so much JMan. I will start checking the individual components of BA185 to see if I can determine where I am losing voltage. Does that sound like a reasonable troubleshooting plan?

 

[JMan]

Yes it does!

 

[MoJo]

Yes it does!

I used at -6db 1kHz sine wave at XLR In; that equals 2.53VAC at XLR Pin 2.

Compress switch is Down, TRX IN Connector Pin 2 = 2.54VAC and Pin 5 = 2.35VAC.

Compress switch is UP, TRX IN Connector Pin 2 = 0.08VAC and Pin 5 = 0.08VAC.

At BA185 card connector Pin U = 1.02VAC and Pin V = 1.01VAC.

Then I started checking the resistors according to the BA185 schematic, but I don't know at what point I should be testing in VDC or VAC.

I don't want to confuse people on this site that are trying to help me by giving them the wrong values.

So my question is, is the BA185 card all DC voltage? If so, where did it change from AC?

 

[JMan]

I think you should have AC on the BA185 board as well. The transformer center tap (Pin "P") appears to be our ground reference if my tired eyes are reading the schematic right.

Since you mentioned VAC readings on Pin U and Pin V separately, I might ask how you are taking your readings. Are you probing these pins individually, with your other probe at a ground point? That's not wrong, it just affects how we interpret voltages. If the transformer center tap is indeed tied to ground, then ~1VAC at Pin U and ~1VAC at Pin V when measured individually, referenced to ground might appear different compared to ~2VAC across the transformer secondary Hi and Lo (measured between Pins U and V), but they would in fact be the same thing.

It seems to make sense that when you have the Compress switch up (bypass?), we wouldn't see AC voltage on this card. Maybe take your meter and probe your way through the card circuit using the schematic. You might not understand every single thing you measure, but it will help you start to wrap your head around what should be happening and whether or not that's what is happening.

 

[MoJo]

This is really becoming fascinating JMan! Tell me if I am coming to the proper conclusions based on what you said in your last post.

The AC signal is coming from the XLR In connection, therefore is no signal or the compressor is "off", then I would not see any AC signal.

There will always be DC signal because the power supply is always pushing 24VDC to the BA185 board.

Is this correct?

 

[MoJo]

UNDOS and JMan,

I am beginning to believe that UP is compression Off like UNDOS thought. Being logical and a great believer in standards, I just assumed that flipping it up (like a house light switch, in the US anyway) means On.

Thanks for being so patient with me. All of my previous projects (50+) were never this difficult!

 

[JMan]

This is really becoming fascinating JMan! Tell me if I am coming to the proper conclusions based on what you said in your last post.

The AC signal is coming from the XLR In connection, therefore is no signal or the compressor is "off", then I would not see any AC signal.

There will always be DC signal because the power supply is always pushing 24VDC to the BA185 board.

Is this correct?

That is correct, yes. The AC signal is the audio signal coming in from the XLR, which either does or does not pass through the transformer depending on whether it is switched in or bypassed. I looked at the routing around the relays, this unit employs a hard bypass which means when it is not engaged, the signal goes straight from XLR IN to XLR OUT, never seeing a transformer or any other component besides the relays - I believe this is the case with the original as well.

The main source of DC will be your power supply, though it may not always be the full 24VDC at every point that you measure it. All of your transistors should have DC flowing through them, so measuring that is one good way to determine if those transistors are still working correctly.

 

[MoJo]

I have completed recording the values of each component; nothing appears wrong to me, but then...
Here is a picture of the BA185 board with the values recorded (you can see the values if you zoom in).

 

[ruffrecords]

UNDOS and JMan,

I am beginning to believe that UP is compression Off like UNDOS thought. Being logical and a great believer in standards, I just assumed that flipping it up (like a house light switch, in the US anyway) means On.

It is a British design. For us, down has always been ON (and still is).

Cheers

Ian