Any Echoplex Experts?

[HotBluePlates]

Looking for any Echoplex Experts, or folks who know solid-state well enough to know good sounding subs for the TIS98 BJT.

I acquired an EP-3 this weekend.  Got a great deal on it, probably because despite its outstanding condition it had a huge amount of noise.  My EP-3 matches in the 3rd Series schematic, although I did see a couple very minor deviations.

I've already checked tape path cleanliness.  The previous owner must have cleaned it right before selling, as it was immaculate aside from a very little bit of oxide shed on some surfaces away from the actual heads, rollers, etc.  I also isolated the (at least first) noise to tackle as being on the playback board, by listening with no input and the tape taken off the EP-3.  The playback board consists of Q1, Q2 and associated components through to the Output Jack, minus the outputs from the wipers of the Echo Volume, Echo Sustain and SOS Record pots.

The noise was also isolated to the playback board by turning the Echo Volume all the way down (wiper to Point 10, grounding output from the playback board).  Dry signal path was very quiet, though I'll need to get an isolation transformer to go between the EP-3 output and amp to break the ground loop.

Ultimately, I found Q2 to be super-noisy.  Q1 showed 2.1mV a.c. on base, emitter & collector.  My meter's leads tend to register closer to 4mV a.c. when not connected to anything.  Q2's base had 2.1mV a.c. (due to direct coupling from Q1's collector), but a highly-varying 28-41mV a.c. at the collector.  So something like 13dB of noise just from Q2.

I wish I could say I found the noise efficiently... Instead I swapped essentially all the resistors on that playback board, one at a time, then re-checked for noise with the EP-3 operating into an amp.  I also cleaned off loads of flux residue from the bottom of the board.  Electrolytic caps are on their way, but have not been replaced yet.

I do have some spare TIS98's (Q1-4) and TIS58's (Q5, Dry path) on the way in case there's some "secret sauce" in using those.  However, I wonder about the risk of the new ones being noisy as well.

Anyone know good subs for the TIS98 that are quiet?  TIS98 Datasheet  I see some suggestions of "use anything with similar hFE" but I know nothing about practical transistor use so I thought I'd ask.

[HotBluePlates]

Got the new transistors in.  Swapped Q2 for one of the new TIS98's.  Success!! Very quiet with just a hint of rushing noise with the Echo Volume turned up to pass only the Playback circuit output.

Then I got greedy... I thought, "I bet I can get rid of the rest of that noise if I also swap Q1."  Q1 swapped, noise increased.  But I had to cut the old Q1's legs to unsolder the old transistor, so no putting it back in.

Lesson learned that you'd like to be able to test the transistors for noise before installation, since they seem to dominate the circuit noise.  I've got plenty of TIS98's on hand now, so I ordered to micro jacks to install on the board to allow the transistor to simply be plugged in.  I should be able to quickly zero-in on a quiet one for Q1 that way, without wasting any of them.

[Willabe]

 

[HotBluePlates]

So I installed the micro-jacks, and sorted through 24 TIS98's to find the quieter ones.  7 of them had reasonably low-noise when tested in the playback board of the EP-3.  So now the Playback side is at least reasonably quiet.

So now I test the Record side, and have a bunch of noise.  I installed the micro-jacks on the Record board, and installed two of my sorted quieter TIS98's, still kinda noisy.  And the signal on the tape seems very, very strong.  I don't need to set the Echo Volume halfway up to get equal volume between dry & repeats...  More like barely above off.

I've already gone through and replaced all electrolytic caps in the EP-3, so now I'm wondering if I'm gonna need to swap resistors on the record board to tame some of the noise.  Or if it's noise from the bias circuit/bias trap.  I'm kinda stumped at this point, as my scope is in another state so I'm not sure when I'll be able to set up the record level/bias as noted in the manual.

The flipside is I can get a reasonably quiet setup with my "wrong adjustment" because the straight-through dry signal (through the JFET buffer) is very quiet.  That allows the Echo Volume to be set for mostly that quiet buffer, while the too-string repeats & noise get turned down.  Still, it would be nice to have it work as it should, yet still with a minimum of noise...

[HotBluePlates]

Continued to tinker...

I noticed the Bias Trap doesn't null out the bias entirely at the collector of Q4.  I've adjusted the trimmer cap for the lowest voltage at Q4's collector, about 200-some millivolts.  That's with 4-point-something volts RMS of bias at Point 3, instead of the 37v peak-to-peak (about 13v RMS) the schematic shows.

Someone on another forum said flippantly, "It's just the tape."

At this point I'm wondering if I should chase nulling out the bias voltage more, or if I've got the bias too low (doesn't seem to improve noise when it's increased at the Bias Adj trimpot), or if I need to be chasing pre-emphasis/de-emphasis.  On the last point, I did notice the 18kΩ resistor from Q2's collector to Q1's emitter resistor string does measure ~20kΩ.  Maybe no big deal, but I imagine that would reduce the de-emphasis a tad.

Also measured a.c. volts at every collector.  I get 2-9mV RMS everywhere except Q4's collector, with no tape on the machine.  So it looks like selecting the TIS98's for low noise worked.  I repeated with the tape in place and seemed to get similar readings.

I also tried measuring the a.c. volts at the Output jack.  Only ~9-10mV with the tape in place, or ~4mV without tape.  Now I'm just confused about whether I'm expecting too much (for low noise), or if I should clean the jacks or what.  My meter does show the a.c. volts are up in the frequency range of the bias (~71kHz), where it gives random frequency readings at other places.

I guess I'll have to ponder & tinker some more...

[sluckey]

Tape can certainly be a factor. Have you ruled out the tape? Heads are also high on the suspect list. They do go bad and low output/noise is often the symptom. I don't know how to test the heads other than substitution. Could be hard to find a replacement.

I know you have cleaned the tape path and assume that also means clean heads? Have you also demagnetized the heads? Usually dirty or magnetized heads will produce a low, dull sounding (loss of highs) output.

[HotBluePlates]

Tape can certainly be a factor. Have you ruled out the tape? Heads are also high on the suspect list. They do go bad and low output/noise is often the symptom. I don't know how to test the heads other than substitution. Could be hard to find a replacement. ...

When I got this EP-3 it came with a Fulltone cartidge (of unknown age).  I did get a couple new cartridges since then, and the noise doesn't diminish when using new tape.

The heads have been something I haven't been able to rule out, and do appear to have quite a lot of wear (this EP-3 is only 42 years old... and I have a lot of wear, too!).  Fortunately, I've located a place to get the exact original heads in unused condition (a magnetics facility that took over the stock of the original head manufacturer).

... I know you have cleaned the tape path and assume that also means clean heads? Have you also demagnetized the heads? Usually dirty or magnetized heads will produce a low, dull sounding (loss of highs) output.

The first thing I did when I got this was clean it!  Surprisingly, the seller must have cleaned it thoroughly first, as there was almost zero oxide on the heads, capstan or even the surface below the tape path.

I haven't demagnetized yet, as I still need to pick one of those up.  I seem to be getting acceptable highs, and have noticed the circuit shapes the sound a lot.  The later EP-3 circuit (which this was when I obtained it) is much darker than the first-version EP-3.  I get much of the guitar range, though the very top end appears attenuated by the Playback circuit (the .003µF and 18kΩ, for example), but the bigger effect is that repeats are compressed.  I guess that's tape for you...

(No, I didn't notice tape compression when recording on a Studer 24-track, but the EP-3 isn't running 30 IPS, I didn't slam the machine when using it, and the Studer costs juuuuust a little more than an EP-3  )