SE EL34 Plexi-verb

[JazzGuitarGimp]

When I first started working with opamps, I had a heck of a time getting them to work properly. So hang in there, because it does get better.

There is nothing inherently wrong with the opamp circuit in the schematic. But understand that running an audio opamp on a single-ended supply, the output of the opamps will sit at 1/2Vcc, or in your case, +7.5V. This is why all the coupling capacitors are required to make it work. It's good you haven't installed the brick yet, as I think there may some changes that will need to be made to your circuit. But without the brick installed, your circuir board should not be getting in the way of the amp's proper operation.

The first thing I would do is measure at the wiper of the Reverb Level pot, R9 for DC voltage. Measure it with the pot fully CW and fully CCW. If you don't see any DC voltage there, then ground the wiper of R9 and see if the amp returns to normal operation. If it does not, remove the reverb board from the amp, and try grounding the reverb return input on the amp (bottom end of R16) and see if this causes the amp to malfunction.

Aaron's circuit would be a better approach, but that circuit requires a bipolar power supply. So we would need to make a virtual ground at 7.5V, then add coupling capacitors to all of the opamp inputs and outputs.

[John_P_WI]

1st stage of the TL072 is set up as a buffer. The second stage is set as an inverting amp with a gain of approx 7. With a gain of 7 it needs the 15 v rail for head room. Note the symmetrical pull up and pull down rails acting as a virtual earth centering the signal in between the +15v rail and ground. Observe the need for dc blocking caps.

Edit.. you guys posted before me, above is in reference to Phils post. Doing this on a phone sucks... can we ever get this place.unlisted?

[Phil_S]

Aaron has posted the schematic supplied by Belton. Please read frankleslie's comments at the link posted above. It is the reason I took his schematic over the one supplied by Belton.

I am not clear on how this works, so it may take some effort to get me there. I am totally willing to do a rebuilt of the board.

I am having doubts here. Am I supposed to be using two op amp chips? I had understood there are two op amps in the one package.

Please excuse me for sounding so stupid.

I'll check for DC on the pot, probably tomorrow, and let you know what I find.

[John_P_WI]

Phil, it is a dual op amp, pins 1 2 3 and 5 6 7 are respective op amps. Pin 7 is the out. I'd probably strap a 100n film cap across pins 4 and 8. Make sure there is a dc blocking signal cap on the output pin 7.

[JazzGuitarGimp]

The TL072 opamp is in-fact a dual opamp, so you've got both opamps together in one package.

And John is right, the second stage has a gain of 7 (33K / 4.7K), and this is actually more gain than you can use here. Assuming the brick can produce a 5Vp-p output, you would need a 35V (5Vp-p x 7) to be able to use all that gain. But keep in mind the TL072, while a good choice for this circuit, does not have a rail to rail output. You will lose about 1.5V off of each rail at the output. So, the rails in this circuit are +/-7.5V, subtract 1.5V for each rail because that's all the closer the TO072 can get at its output, and you have a max undistorted signal output of (7.5 - 1.5) x 2 = 12Vp-p. With 5Vp-p coming in, the gain need not be any more than 12Vp-p / 5Vp-p = a gain of 2.4. I would replace the 4.7K resistor with 13K. But I think there are more important things in the design that need to be addressed. I need to scrounge up the datasheet for the brick before I can make more suggestions.

[martin manning]

The circuit you linked from Frankleslie is using BTDR-1, and you are using BTDR-3, right? If so you should use the mono version in the Belton BTDR-3 application notes. Since you have a single 15V supply, the terminals marked Common should be supplied with Vcc/2 from a voltage divider from your 15V to ground, just like Frank's circuit. I'm not sure why he chose to use two separate voltage dividers (maybe some one else can comment on that) but it will work in the Belton app circuit. Re your question on the TL072 pinout, pin 4 -Vcc goes to ground; pin 8 +Vcc is the 15V supply.

[John_P_WI]

The circuit you linked from Frankleslie is using BTDR-1, and you are using BTDR-3, right? If so you should use the mono version in the Belton BTDR-3 application notes. Since you have a single 15V supply, the terminals marked Common should be supplied with Vcc/2 from a voltage divider from your 15V to ground, just like Frank's circuit. I'm not sure why he chose to use two separate voltage dividers (maybe some one else can comment on that) but it will work in the Belton app circuit. Re your question on the TL072 pinout, pin 4 -Vcc goes to ground; pin 7 +Vcc is your 15V supply.

Martin, pins 4 and 8 are the rails... right?

[martin manning]

Yes, sorry, the pinout diagram is posted above.

Also note the Belton mono circuit includes the parallel path, so that network can be eliminated from the amp schematic.

[Phil_S]

This is going to take me a while to digest. Here is the data sheet for the brick. Yes, I am using BTDR-3.

I gather I did OK using one chip here for the 2 op amps needed.

My question on Pin 4 is whether the 10K/10K divider gets connected there; since Pin 4 goes to ground, does it really matter if I just take the tail of the divider and simply ground it or does it even matter? Ground is ground, yes or no?

When you say "Vcc/2" I understand that to mean the 15VDC supply divided by two. IOW, there are two dividers, one is a pair of 2M resistors and the other is a pair of 10K resistors.

There are 3 terminals marked Common on the Belton circuit. Are you telling me all 3 need to be fed half of the 15V supply and that is done by using a voltage divider?

I'm going to copy a couple of Frank's comments here as a convenience. I believe this explains his thinking on 2 separate dividers.

I have installed a Belton reverb brick in a newbuild and it Works very well. It taps off the signal from the volume pots wiper and inserts it back into the second grid of the phase inverter. The Associated Components is a high impedance buffer (dual op-amp) and an amp that gives 5 to 10 times gain, mine is about 7. The op-amp is a TL072 but any low noise Device will work. The Belton distorts (nastily) if the signal input is above 1,5V RMS so a bit of tweaking has to be done. I enclose a diagram of the one I found to work well in my build.

The two 2M2 resistors are for setting up bias to the first op-amp and the two 10k are for the second. The first op-amp is a buffer with a gain of one. Check out the full schematic for PSU details.

Also note the Belton mono circuit includes the parallel path, so that network can be eliminated from the amp schematic.

Martin: what network? I don't follow ;-{

Please be brutally honest with me here. Should I be rebuilding the board? Parts are cheap. I am totally willing, but I just don't want to find myself going around in a circle.

Many thanks to everyone for the help.

Phil

[xtian]

Here is a schemo for the BDTR-2, for comparison. http://byocelectronics.com/reverbinstructions.pdf

[JazzGuitarGimp]

Phil,

This would be my approach. There are five trim pots on the board - I feel that once these are set, they won't need to be adjusted again. They are:

RT3 - this adjusts the signal level coming from the amp. The brick only accepts signals up to 3Vp-p, so you're going to need to attenuate the signal coming in.

RT1 and RT2 - these adjust the amount of "depth" for each reverb channel. My guess is they will want to be all the way up, but I am not familiar with the Belton brick, so better safe than sorry.

RT4 and RT5 - adjust the signal level coming out the two reverb brick channels, before they go into the mixer (U1B). These too, will likely want to be all the way up, but again, not know the brick, I cannot say for sure. I have set the mixer gain to about 5.5, which is based on data from the Belton datasheet.

Potentiometer R6 is the only panel mounted pot you should need for Reverb Level.

I have made a virtual ground with resistors R1 & R2, and capacitor C2. These two resistors should be of the 1% metal film variety. This circuit produces the +7.5V reference to make the opamp think it's running on a bipolar supply.

The first half of U1 (TL072) is a unity gain buffer.I am thinking you will have way more signal than you need at the input, so there is no reason to make this device a gain stage.

The second half of U1 is a two input mixer, used to mix the left and right outputs from the brick. It has a gain of about 5.5, which should be just enough to get the maximum, undistorted signal level from the brick up to the maximum, undistorted signal that can be produced by the opamp, and this is why I suspect trimmers RT4 and RT5 will be fine all the way up. In fact, you could probably build the circuit without RT4 and RT5 and not miss them, though they may prove nice in terms of "blending the two outputs to taste."

Let me know if you have any questions.

[JazzGuitarGimp]

One other point I should make: The inputs of the opamp, especially the inverting inputs (pins 2 and 6) are susceptible to RFI (think tube grid), so keep leads into these inputs as short as possible. I have included a small cap in the feedback loop of U1B. It has a -3dB point at 10KHz, and it is there to keep the opamp from going into oscillation. U1A does not need this as it is unity gain. And U1B only questionably needs it as its gain is relatively low, but it's good engineering practice to include it.

I also wanted to mention that the BTDR-3 datasheet doesn't say anything about this, but I am making an educated guestimate that it is not a good idea to parallel the depth pot terminals 1A & 2A - and - 1B & 2B, and this is why I have used separate trim pots here.

I may post more thoughts as time allows, but right now I've got a wife coming home from a 10-day trip and I've got me some CLEANING to do!

Feel free to shoot over any questions.

Lou

[martin manning]

Also note the Belton mono circuit includes the parallel path, so that network can be eliminated from the amp schematic.

Martin: what network? I don't follow ;-{

In your whole-amp schematic R14, 15, and 16. R14 is the parallel path. In the Belton diagram the connection from the input to the + input of the output amp replaces that function. Lou has provided a circuit that eliminates that, and I think that is probably a better way to go.

I also see in the Belton diagram that there is a large 220u electrolytic at the regulator output of the 5V supply. You might consider making that change.

[Phil_S]

Lou: You are very fortunate. She leaves you alone for a while AND then she returns. It is the best of both worlds! Thank you very much for being so thoughtful and generous with your time, to create a circuit for me to use. I can see now just how lost in the weeds I am with this. My assessment is that it would take me months to do enough reading to get up to speed on what's actually required. So, I guess you must know how much I appreciate that you threw me a fish.

On the depth pots (not being argumentative, just exploring):
I understand "depth" to be the delay. This is a fixed interval on the earlier BTDR-1 and BTDR-2. Choices are 2s, 2.5s, and 2.85s. There is room in the early designs only for a Level pot. BTDR-1 is mono only. BTDR-2 adds stereo. BTDR-3, the most recent in the line, allows for adjustable delay. So, this is supposed to be the equivalent of short, medium, and long spring tanks in one package. On this basis, I got the idea that a panel controlled pot is intended, further suggested by Belton indicating a dual pot. Besides, I've already got a hole in the front panel. I suppose I can leave a dummy of some sort in there if I don't use it.

On the "level" control, I understand this to be the thing that controls the amount of reverb (for lack of a better expression). I can better understand your design here, which is to simply turn it all the way up (if that works out sonically) or find the sweet spot and leave it there. Then level control is managed by the A100K pot. In that regard, can I just skip the B10K pots, and just use the resistors as a fixed mixer? I suppose the problem there is that I need something in place for tweaking?

I'm not sure what all the questions are on depth and level ;-}

Martin: On the parallel network:
Do I remove R15 and R16, and simply send and return to each side of R14?

Or do I remove all three resistors, break the circuit continuity at that point, take the output from the Treble pot for send and simply return to C10, which is input to the recovery stage?

What are the implications if I keep the network in place?

My thinking on the parallel network (I knew what it was, just didn't know what it's called ;-} ) was that the BTDR-3 has no OFF switch. I figured if I turn it all the way down so there is little or no output, the network serves as a bypass of sorts. Are you saying this isn't needed? I have some concerns about how to pass the signal through when I don't want reverb. I don't understand/don't see the parallel path on the schematic drawn by Lou. All I can see is that the signal has to go through the brick.

Do I want to install some sort of switch to kill the reverb and direct the signal around it? Or am I just barking up a tree here?

Oh, and all this is really terrific. I'm learning a great deal. Thanks.

Phil

[martin manning]

Sorry I guess I wasn't clear. Lou's circuit is like the one you had previously so you would keep the R14, 15, 16 network as you have it to let the dry signal pass through and mix with the wet before going into the following stage.