Echoplex EP-2 Erase Bias

[SteveBAnalog]

Hello All. Thus far in my troubleshooting my echoplex, this thread: http://tapeproject.com/smf/index.php?topic=1100.0 on this forum had the most valuable information.. so I thought I ask a related question to the techs here. Thanks so much for your time. Any advice or links will be gratefully received.

I own an echoplex EP-2. When I purchased my EP-2, it had very weak repeats. I replaced most of the caps. The filter caps were replaced by the previous owner with a higher value multican 40/20/80/80. The voltages appear within spec . I have tried different tapes and physical adjustment to the heads. Demagged and cleaned numerous times.

I traced the problem to the oscillator circuit. Replacing the .047 c15 with a fresh .01 and setting the R26 trimmer to very low resistance gives me strong full echoes, repeats and self oscillation when driven. The unit doesn't fully erase. When the black lead from the erase circuit is scoped I find a distorted sine wave which looks to be 25 khz. (full disclosure, I am a new to my oscilloscope but I think I got the math right) This seems low for ac bias. The Orange lead to the record head shows the same frequency but smooth and a much higher amplitude with the r26 trim lowered. I don't totally understand how the bias choke and capacitor work. I tried adding a variable capacitor, c17 but I don't register a change at V2B per the schematic instructions. Do you have any recommendations for further tests? How might I raise the ac bias for the erase head?

I realize I am a little out of my league on this circuit but, I feel like I am so close and not ready to give up just yet. Especially when i got the echoes to sound amazing... I just want deeper erasure!

thanks so much for your time.

[stellavox]

Looking at the schematic, you might want to check the resistance of the erase head to see if it is around the 28 ohms given - can measure it "in circuit", right across the head/connections as it is coupled to the oscillator thru an .01 cap.  If that cap isn't shorted/leaky, it should NOT pass DC. 

I'm confused a bit by note 3. The schematic shows R26 to be a 33K fixed resistor: but the note says it can be replaced by a parallel coil/cap in series with a 220K trimmer.  The coil/cap is a record "bias trap", a circuit that is "supposed to" be tuned to the bias frequency and "trap" it so it can't make it into the record circuitry (V2) and possibly cause intermodulation distortion.  I'm troubled by the 220K value which seems WAY too high, considering the fixed resistor it replaces was 33K.  I'd opine that it is a misprint and should be around 25K or so - which would also make sense since you are setting the current trimmer "lowered".  The erase level is "set" by the .01 cap C15.  Try replacing it - suggest 250VDC MINIMUM and see if it may be the problem.  If not, you could try increasing the value of C15 to .015 or.022.  However,  maybe the circuit is working "as designed" or is "as good as it can be".  The erase oscillators on pro tape recorders are supposed to generate a really pure sine wave - distortion adds "noise" to the recording.  Realize that this is a "much cheaper" circuit.

Charles

[SteveBAnalog]

Thanks for your help Charles. Here is a little more info:

1.The erase head measures 21.8 ohms so that seems ok.

2. The R26 value definitely isn't a typo as it appears across multiple schematics of different serial numbers (these units are interesting in that each serial number grouping is a little different, different voltages, varying values and mine has proven to have a few "in between" design attributes, including some values more similar to the EP-1.) R26, along with the 18mh choke and the variable capacitor c17, comprise the bias trap circuit. Some schematics include that in the circuit proper, others footnote to it as a possible improvement. I don't understand c17's role and didn't see a change when replacing it with a variable capacitor.Also when I trim R26 down to nothing, I get the loudest echoes and the most "erasure.." though perhaps I am just recording over at a hotter level? As I increase resistance, the echoes are still pretty good up to a point but the previously recorded material becomes louder. Originally my c15 was .047 which corresponds to an earlier schematic... but this value didn't induce multiple repeats or oscillation. Changing it to .01 produced repeats, the original problem I was trying to solve.

3. R27 varies across all the schematics. Currently mine is 4k7 which increases the plate voltage 220v on the oscillator tube... when reduced to 560 ohms as in some schematics, the plate value reduces to 160. Both values correspond to iterations of the schematic, but changes here didn't seem to improve erasure.

4. c16 is supposed to set the resonant tank. When increased to .002 from .001 I noticed the slightest increase in bias frequency.

5. The oscillator tube is new, but I have no way of testing it. I have swapped the old tube back and noted the same results.

Again as a newbie and amateur here, I very much appreciate your time reading these and helping me troubleshoot. I am grateful to learn from you all. Cheers.

Stephen

[stellavox]

The purpose of C17 is to "tune" the bias trap to the oscillator frequency, thereby minimizing the amount of bias voltage "leaking" into the record amp V2.  The way to adjust it would be to put a scope on the plate (pin7) and, with no audio signal, adjust the cap for MINIMUM bias "bleed thru".

R26 adjusts the amount of bias added to the record signal.  "In theory", you "optimize" the bias by recording a HF audio signal, say 10Khz or so and while monitoring it's Playback level, increase the bias signal till the 10K peaks, then increase the bias a little more till the signal falls by a few dB - maybe 2 to 4.   

[SteveBAnalog]

Thanks again for the info. I did some more work and testing. Here are my results:

Original c15 bias cap at Erase was .047. This corresponds to one of the schematics, though most schematics use a .01 here. When I switch back to the .047, the ac bias signal is now a pure sine wave but increases to way over spec, in the area of 200kHz. Also with the .047 cap, I don't get repeats or much echo at all, and no erasure. .022 is the same story.

One tech who helped me via email also suggested that the record/erase circuit is very sensitive to impedances and to use low esr caps, ceramics where often in the original. I might order a few different values to test this out.

I did come up with a temporary solution:

1. the bias trap doesn't seem to be working and doesn't have an adjustable cap originally but a 240pf. The variable cap I tried have seems to make no change. I have removed the trap for now leaving R26 as a trim and adding a 47k in between the orange record head lead and c13 per the original non-bias trap design.

2. I added a 10k trim at R27. lowered to around 385ohms, this gives me the best amplitude on the erase signal. It also lowers the plate voltage on the oscillator tube to 156v. The schematics vary on this value, some 219v others 125v

2. c15 is back to .01, but I also added 2 .001s which I had on hand. So the .012 cap gives me a much better window as I trim R26 for maximum echoes to erasure ratio. Note, with this value, my bias signal is still in the 20-25kHz range. It should be 50-70kHz.

With these adjustments, the unit is basically there... sounds really good, and erasure is at about 90% I'd say. The trade-off is that for this level of erasure, I don't get runaway repeats. If I increase R26 I can achieve the runaways, but the ghosts of recorded sounds on the tape come back.

And I doubt I can put this to rest knowing that things are still not quite right and I still am after the circuit that gives me correct bias and a working trap.


I am wondering:

I notice that the leads to the erase/record head are wrapped in the same shield unlike factory where the erase and record leads were in separate shields. The head measures correct and looks original, but perhaps has been messed with. Could the the leads sharing the same shield wrapping influence the bias frequency issue?

Also, the one area I haven't addressed is that the multican filter caps are non original and a few of the values are different. ie.
c18c is 100uf, c18d is 80uf, c18b is 20uf but was added separate to the multicap.  Could any of these cause a weird interaction in terms of the ac signal?

Again if any of these diagnostics rings a bell, I'd love to hear back. If not, no worries, I know I am asking a lot..but hopefully the troubleshooting will be helpful to others too. Thanks so much for the knowledge you have shared already.