Replacing & interchanging reverb tanks

[Mango Moon]

I've got an amp in a head cab that came factory-equipped with an Accutronics 8AB2A1B reverb tank, which is a short (9"), three-spring, medium-decay model with grounded input, and I'd like to replace it with a longer (17") three-spring model with long-decay, which would be the equivalent of a 9AB3A1B. However, based upon my research, no such reverb tank exists, not now or apparently at any time in the past, not from Belton, Accutronics, Ruby, MOD or anyone else. The only thing that comes close is a 9AB3C1B, which would cover all the required bases, except that it would have an insulated input, rather than a grounded one.

Now, when I ask around about this, I'm told that the insulated specification for the 9AB3C1B will make it incompatible, because it would be subject to excessive radio frequency interference (RFI), and that to make it compatible, I would need to modify it somehow (i.e., solder a ground wire to something, somewhere?). However, the builder of the amp tells me that the insulated input version (the 9AB3C1B) . . . "will work just fine - we don't care about grounded input because the amp ties the grounds together anyway".

Needless to say, I'm inclined to trust the builder of the amp over the average respondent on a public board, but nevertheless, it troubles me that I'm seeing these two different camps or schools of thought on this, and I definitely don't want to introduce any potential for RFI to the amp. Can anyone shed any further light on this subject, any clarification?

[olddawg]

"Potential" RF interference will not hurt your amp. Worst case scenario would be you might hear an AM radio station once in a while. If the jacks shields are tied to together in the immediate circuits it will be fine as the manufacturer says. If you do have a problem just clip a lead from the tank body to the jack shield. If it solves the problem solder a wire to the same points. I doubt it will be necessary.

[Mango Moon]

Olddawg, many thanks for your reply.

I strongly suspect that you and the builder are both right on this (i.e., that RFI won't prove to be a problem), but I've learned over the years that its never a bad thing to be absolutely certain, before one actually gets started with purchases and so forth. Also, if it does in-fact prove to be a problem down the road, the solution you've proposed (soldering a wire from the body to the "shield") is precisely what I've seen referenced elsewhere, but I'm afraid these instructions aren't particularly clear to me. I understand that one end of this newly introduced ground wire would be secured (soldered) to the inside wall of the tank body itself, somewhere near the RCA Input jack, and that's no problem, but what and where is the "shield"? In other words, where would the other end of that ground wire be affixed to? Are we talking about the soldered connection for the shielded part of the RCA connector on the perimeter of the inside insulator?

The other thing I'd really like to address is the diagnosis of an existing Accutronics 9AB3C1B tank that I have here. I already own a U.S.-made Accutronics 9AB3C1B tank that I purchased new a few months back, but it doesn't seem to work. I see nothing overtly wrong with its springs, solder points or any of its other internal parts, but when I use a multimeter to try and determine its integrity, I'm not getting the readings that others have said I should be getting. Perhaps I should start another thread to discuss this topic (?), but suffice it to say, with the tank sitting here on my desk, I'm not getting the expected 8-10 ohms at the input or the expected 2575 ohms at the output. Hence, I'm having trouble determining whether the tank is any good, or whether it may have arrived here months ago, dead-on-arrival.

[g1]

Agree with olddawg, the builder has probably tried it both ways and found no difference. I have tried both ways in some amps and found no change.
That being said, if you use the MOD tanks, they have jumpers inside so you can quickly lift or connect the ground at either end.

[g1]

As far as the grounded input goes, the outer connection of the RCA jack gets connected to the body of the tank. If the input is not grounded, that means the RCA ground terminal (outer connection) is insulated from the body of the tank. So your jumper wire would go from the RCA ground terminal to the tank body. It's a little confusing as the tank input can be called "ungrounded" yet it can still be connected to the amp's circuit ground.
On your other bad reverb tank, what resistance readings are you getting?

[Mango Moon]

O.K., so if I understand you right, I'd simply be providing a ground jumper from the ground at the backside of the terminal to the chassis or metal body. Good enough! I can handle that.

As for the readings on this 9AB3C1B that I have here, its a real puzzle. With the multimeter set to 2K ohms, I'm getting a consistent reading of 0.225, which I'm assuming is 2250 ohms (even though that conversion doesn't seem to make sense?), and if that's the case, then the Output side of things appears to be good. However, on the Input side, when I jumper across the two terminals with the multimeter set to the lowest possible setting (+), I'm getting something like 0.001 which makes no sense. I'm beginning to wonder though, if this is perhaps a matter of personal misinterpretation on my part, because the only settings available on my current multimeter are 2M, 200k, 20k, 2k, 200 and +, and maybe this multimeter is simply incapable of reading a resistance that is that low?

[g1]

All tanks you mentioned have 2nd and 3rd digit AB so they have the same input and output impedance and should measure the same. Spring Reverb Tanks Explained and Compared | Amplified Parts
10ohm input impedance should measure about 1ohm resistance, 2575ohm output impedance should measure about 215ohm resistance on your meter.
See what your 8AB working tank measures. I think .225 would mean 225ohm on your 2K range, but the .001 on the 200ohm range doesn't sound right. I would expect a 1ohm reading on your 200 range to read as 1. See what the 8AB tank reads.

Edit: I realized the (+) setting is probably diode check so .001 for 1 ohm is probably correct. You will probably get a 1 reading on the 200ohm setting.

[Mango Moon]

Unfortunately, I don't have the original (9") 8AB2A1B tank readily available to me at the moment, but I can check it in the morning to confirm and then report back. In the meantime, the Output reading on the larger 9AB3C1B is indeed 225 ohm when set at 2K range (I double-checked it), so that now makes sense to me given the contents of the Amplified Parts article and the fact that it should be expected to read 215 ohm or so. Also, I confirmed that the Input reading is in-fact 0.001 when set at the lowest "->+" ohm setting (whatever that setting means?), and yes, it reads 1.0 when set at 200 ohm. Accordingly, it would appear that this 9AB3C1B tank is mechanically and electronically functional, so that's some good news!

However, I must say, I sure got an education on this one! I thought all along that if the Output impedance is supposed to be 2575 ohm, that it should read "2575 ohm" on the DMM, not 215 ohm? Same with the Input impedance. If its supposed to be 10 ohm, who would expect a reading of 1 ohm to be correct? After all, the units of measure on the faceplate of the DMM don't say anything about DC resistance - it says Ohms? I don't know about others, but that's a tad confusing to this novice.

Anyway, you've been of great help, and you've helped me answer the central question with confidence, effectively saving me about $40 (when including shipping costs) for a new reverb pan! Many thanks!

[g1]

To try and be brief, impedance and resistance are not the same although they are both measured in ohms. Impedance is how something responds to AC, resistance is to DC.
A meter or DMM will only measure DC resistance, not impedance.
Sometimes the two figures are close, like when a speaker with 8 ohm impedance measures around 5 ohms resistance with a meter. Other times they can be quite different, like the reverb tank example where the impedance (at 1khz) is 10 times the DC resistance.
Have you tried the other tank you have yet? If the readings are good and the springs are all connected but it does not work, then you may just have a connector problem.

[52 Bill]

Agree with all of the above information, the only difference in the tank is the input connector type. Just add a ground wire to the input if it needs one and you should be good to go.

The main reason that there are ungrounded inputs is to eliminate the ground loop when both end grounds are connected. Fender uses shielded cables for both input and output cables, so they use an ungrounded input jack. Mesa uses a grounded input jack, but the drive line is an unshielded single wire.

[Mango Moon]

I want to thank all of you guys again, but especially g-one and olddawg for your assistance on this. I retrieved the original (Korean-made) 9" 8AB2A1B this morning and tested it, and it gave me the exact same readings as the USA-made 9AB3C1B that I have here, so it looks like I'm good to go. The amp that I'm hoping to install it in, is out getting a new cab made, so it won't be available again to install and test this 9AB3C1B for a while yet, but when I do, given what I've learned here, I don't anticipate any problems with it. Thanks again!

[Enzo]

Bill, there are also reverb drive circuits that do not have either side of that jack grounded for circuit reasons.. Look at the PV Classic 30 reverb drive as an example. Marshall DSL100. Crate V50. I guess most of the solid state drive by ICs.

[52 Bill]

Bill, there are also reverb drive circuits that do not have either side of that jack grounded for circuit reasons.. Look at the PV Classic 30 reverb drive as an example. Marshall DSL100. Crate V50. I guess most of the solid state drive by ICs.

That's right, I forgot to mention those that drive the tank in the feedback loop of an opamp need the isolated input jacks.