Building a Dundovic tape playback preamplifier

SeleniumFalcon

I'm starting to think about what kind of enclosure I need for this thing. Has anybody found a really cool source for a component enclosure? Besides the usual Bud box?

newbie308

If you draw what you want and don't mind waiting a little bit, a CNC machine shop can make a custom enclosure for you, or you can 3D print one.

SCompRacer

Hammond is similar to Bud but they offered some wood end cap options. For large projects HiFi2000 is my go to. They can machine and anodize after machining. Excellent quality.

For small projects look at Eightwood on Amazon. These are similar to the boxes Welborne used. The PCB slid into slot. These boxes split so you can mount smaller boards and still gain access.


Carl's DAC, HiFi2000 chassis. I used Front Panel Express to design.

SeleniumFalcon

@newbie308 and @SCompRacer thank you for your suggestions (beautiful job on the DAC). I'm toying with adding VU meters and the necessary buffers to reduce the distortion they can cause. Ive measured the voltage gain and at full output it is 33dB which might not be sufficient with the Crown's head's output. I might need to add another gain stage with an emitter follower to keep the output impedance low. Otherwise the output impedance is going to depend upon where the 10k volume pot is set. I am using extremely low capacitance connecting cables but an emitter follower would be good. I can always add a bypass selector to try with and without.
The harmonic distortion is 0.099% at 1.0kHz after carefully balancing the source level versus the output level. It rises very rapidly if the input is raised too high, then there needs to be enough signal leaving the output to match the analyzer's necessary level for the measurement. I also have to turn off the work lights around where I'm working. I'm going to measure the operating voltages to make sure the circuit is biased class A, the collector-base voltage needs to be 1/2 the B+ for that to happen. I may need to add heatsinks for the extra heat.
Now I'm going to build the second board.

newbie308

If you use an aluminum enclosure it will reduce RFI, but you may want to add copper foil to the inside and bond it to ground to block both EMI and RFI.

SeleniumFalcon

Good suggestion, thanks. I have some copper mesh screen that's been in a part's bin for a long time, maybe I'll put it to good use. Matt Polk might have given it to me when I was working on his OTL amps a while ago.
Most commercial tape recorder designers didn't spend much effort or money on shielding the electronics sections from the rest of the machine. My belief is that they felt the leading cause of noise was caused by tape hiss and that any shielding effort was buried under that. Professional tape recorders, on the other hand, did spend some considerable effort on noise shielding. Balanced line connections and four layer connecting cables were the usual efforts.

SeleniumFalcon

I've finished the second circuit board and had to troubleshoot a problem. The second board had an incorrect frequency response and very low output when compared to the first board. It seemed to me that the problem had to be near where the feedback between the two transistors. I started measuring resistances between the two boards and found the problem. I had installed the wrong value resistor at R7, instead of the correct 2.2k ohm value I had used a 221 ohm value, careless mistake.

SeleniumFalcon

I built a platform where I could mount the two circuit boards and the high frequency trim controls and the volume controls. I used pieces of scrap oak board and Masonite and glued everything together. Now I can continue testing both circuits and connect to a tape head output.

SeleniumFalcon

I can connect a signal generator as an input and measure the harmonic distortion on a venerable H-P 334A analyzer powered by a variable DC power supply.

SeleniumFalcon

Clipdat

omgggg.. more of these bokeh overload photos with crazy DOF. **** love these.

SeleniumFalcon

The next step is to connect the Dundovic preamp to the Crown CX800 transport's head assembly and use a calibration tape to adjust the high frequency playback response. This tape has multiple frequency tones all referenced to a 400Hz level of 0dB. Once the level on the preamp is set to 0dB then the following 18 tones should read -10dB, going from 20kHz to 31.5Hz. The results are going to depend on the Crown's playback head characteristics and the preamp's circuitry. Both channels are excellent, within 1.5dB across the full range.

SeleniumFalcon

treitz3

Now, Ken. I'm a little disappointed in you now. Where are the 3D check boxes?

Pffft...and here, I thought you were being thorough...

Tom

SCompRacer

Outstanding work! Pops told me the only people that make mistakes are the ones doing something.

Is that mockup finalized and the chassis size you will use?

SeleniumFalcon

"Where are the 3D check boxes?"

I look at the part on the check list, pick the resistor out of the packet, measure it to make sure it's the right value. Then I use a small piece of Scotch Brite to clean up the leads, solder it in place then check the continuity with the next connections and snip off the excess leads. Then check off the 3D box. Somewhere I saw 2.1k and when I reached for the part I picked up 221 instead. A cognitive disconnect.

I've been spending the afternoon listening and it sounds good, but not blowing me away, yet. There's no graininess but it seems to lack definition. It could use about 10dB of additional gain, I have to turn up my line preamp to "11:00". Channel balance is good and bass definition is fine. There's a veil somewhere, I'm using the battery for power, I'll try the power supply tomorrow.

SeleniumFalcon

My next goal is to have more audio signal available for my preamp. Each calibration tape has a magnetic flux level measured in nano webers per meter (nW/m) which when played back will have a certain level. For example a flux level of 185nW/m tape played on a Revox A77 should produce 0.77V output with the volume control set to maximum. With the Dundovic and the Crown playback head I measured 0.07V with the volume set at around the 2:00 o'clock position using the 185nW/meter level of the tape. So, I need to add an additional stage of gain to increase the output but not add distortion or alter the EQ settings.
At hand I have two choices for experimentation: 1) use the "flat" setting on the Shure M64 that I built to follow the Dundovic's output, 2) use the vacuum tube line stage I built for use with the Advent Dolby 100A.
The first part of the plan is to measure the Dundovic's frequency response with the EQ adjustments to be able to compare with the output after going through both the Shure and the tube unit. Also to be able to use my signal generator as an input source to be able to not use the test tape at this stage. So I need to set the generator's output level to the same level as the Crown's head playing the calibration tape.

SeleniumFalcon

I adjusted the Dundovic's volume level to give 0.77V and made slight readjustments to the treble response to produce a flat response. Now with the Shure M64 as another gain stage my preamp is set to around the 9:00 o'clock setting. Things are starting to sound better, especially harmonic overtone and a sense of where the instruments and the orchestra are in relation to each other. I think I need to leave it powered on for a while to let everything settle.

SeleniumFalcon

I've been collecting 4 track tapes that reviewed as being among the best sounding, most are classical and the majority of those are Barclay Crocker Dolby B encoded. Of the jazz and popular category most are almost impossible to find and when found sell for high amounts. In order to evaluate how this tape preamp sounds I needed a tape that would be revealing of subtle details and give me a way to know how accurate the playback sound was. I borrowed a page from Paul Klipsch and decided that I would make some recordings in my listening area myself. This is what Mr. Klipsch did, made recordings of musicians and used the tape playback to judge how speakers and amplifiers reproduced what he had made. In fact copies of these tape were commercially sold and used as demonstrations of how realistic his speakers were.
I used two Sony ECM 21 microphones, battery powered electret condenser types, fed into a battery powered Sony MX 12 mixer and into the Crown recorder. At first I recorded my wife playing her mandolin with the Sony mics about 12" behind the plane of my two CLS IIa speakers. Then I followed this with her playing a meditation gong using a felt covered wooden striker. Talk about harmonic overtones and long sustain! Now I have a way of hearing a tape and judging how it compares to the actual sound itself.

txcoastal1

Need more cowbell

that's super cool Ken

SeleniumFalcon

This appears to be an Ampex 602 tape recorder.




Digital copies of Mr. Klipsch's recordings can be purchased from:

https://www.highdeftapetransfers.ca/search?q=klipsch

Jetmaker737

Brilliant work. And fascinating.

SeleniumFalcon

This is a link to a very helpful website that has tabulated and rated both tape decks and prerecorded tapes:

https://reel-reel.com

I've been trying to collect tapes rated 9 or 8 points and listening to them on the Crown with the Dundovic/Shure playback stage. These certainly sound great, but what has surprised me is the differences when listening to less highly regarded tapes.

Jetmaker737

SeleniumFalcon wrote: »

This is a link to a very helpful website that has tabulated and rated both tape decks and prerecorded tapes:

https://reel-reel.com

I've been trying to collect tapes rated 9 or 8 points and listening to them on the Crown with the Dundovic/Shure playback stage. These certainly sound great, but what has surprised me is the differences when listening to less highly regarded tapes.

Interesting. I've been collecting tapes for 4 years. I have over 200 tapes. both 2 and 4 track, classical and jazz, 7.5 ips. I take detailed notes for each tape on SQ and performance (I also track prices). I'd never seen this site before. My ratings tend to be a point or two higher in many cases. In some cases we agree. But I did notice that most of the ones I looked at on this site had only one owner (so I assume only one reviewer).

I've definitely had some clunkers for SQ but not as many as I'd expected going into it.

SeleniumFalcon

I think the reviewer has very high standards, neither tape deck or tapes has a score higher than a 9. I just bought a great sounding Columbia tape of Vladimir Horowitz playing Beethoven, Debussy and Chopin. It has amazing dynamic range and clarity with out a hint of distortion. I did notice that the recording engineers had panned the sound of the piano to the right during the famous opening movement of the Pathetique sonata and then the piano sound was balanced more evenly during the following movements. But, other than that it sounded pretty spectacular, certainly much better than any Columbia vinyl disc ever did. That's the beauty of getting the most out of these older 4 track tapes you can hear these amazing pianists and violinists from the golden era who didn't record for Decca/London or EMI with audiophile level quality. I have a second Horowitz from the same time period coming soon. The other nice thing is these tapes aren't very expensive and in fairly good shape.
I bought the London recordings of Wagner's "Tannhauser" and "Siegfried", both rated a 9 on the site and they are pretty spectacular. The dynamic range is tremendous and has some bass information that would cause tracing distortion on a record. The mastering engineers definitely knew that, for the time period, the serious listener would be buying the prerecorded tapes and made an effort to push the limits.
The next time you're playing a 4 track tape see if side B doesn't sound better than side A does, especially the transient detail. I have a theory of why that might happen.

Jetmaker737

This is my experience as well with these old tapes. In general the sound quality is very, very good. Sometimes simply spectacular. Generally superior to my hi-rez FLAC over the PS Audio Directstream Jr. I had read a lot about poor manufacturing processes in dubbing of tapes so expected a high rate of fallout due to SQ. This has not been the case. Most of my tapes are circa late '50's to mid-60's releases. So perhaps that's a factor.

I have not noticed the side A/B disparity. I will listen for it. I have noticed though, that on 2-track tapes especially there can be a bias towards the right or left channel such that the balance has to be adjusted. I have not ruled out that my deck may be the culprit but the bias can be right, left or none at all. When it exists it is more often to the right side than left.

SeleniumFalcon

I am on a Yahoo group that has a considerable number of recording engineers posting. A while ago there was a thread dealing with the best way to duplicate a magnetic tape to minimize transient smearing. Several members agreed that running the source tape backwards when duplicating resulted in a better sounding copy. The reason for this is that the transient waveform, of a cymbal for example, would be presented to the recorder copying it "tail first" instead of "head first". The transient would be gradually building up from the decay to the initial transient attack. This would better prevent ringing and smearing of the transient waveform.
A friend of mine makes home brewed ultra high quality live recordings at his house using highly modified Studer A80 recorders. I posed this theory to him and he asked his Studer technician if there was any validity to this. He replied that yes this was common practice at Abbey Roads studio when he worked there. Also, in an interview with Kevin Gray a well respected mastering engineer he says that he always uses this method when duplicating master tapes.
Here's where my theory comes in. When a 4 track tape is mass duplicated both side A and side B are copied at the same time. The duplicating deck plays all four tracks at once, side A runs as normal, but side B is being simultaneously played and copied running backwards. I believe this results in a slightly better duplication of side B than side A. Of course there are lots of variables, do all duplicating companies do things the same way? Plenty of other factors come into play, such as the tape's content to begin with. Are there lots of transients on the tape to smear or not smear? Anyway, that's my theory.
Have you found that tapes duplicated by Magtec sound better or worse than those duplicated by Ampex?

Jetmaker737

SeleniumFalcon wrote: »

Have you found that tapes duplicated by Magtec sound better or worse than those duplicated by Ampex?

That's not variable I've been tracking but easy enough to retroactively pull it in. I'll need to put that on the to-do list

Is the smearing due to a transient hysteresis effect of some sort? It would make sense then to record from low amplitude to higher, therefore from the decay towards the attack. It's very interesting and I'll make sure to listen for it next time I play a 4-track tape.

SeleniumFalcon

The Magtec tapes tend to sell for a bit more than other duplication firms and the artists are not ones I listen to. With a few exceptions. You can set up a search for just the word Magtec and get lots of responses.
I believe the transient hysteresis aspect is correct, better to build up to the transient peak as gradually as possible. Anyone with a tone burst source and a storage oscilloscope should be able to verify. Maybe I could get a government grant to set up a research lab? Then that Ampex ATR102 would be first on the requisition list.
Can you give a few examples of tapes that you find sound good?

Jetmaker737

Here are some suggestions (no particular order). These are 4-track tapes (assuming that's what you're after). I have many more 2-track than 4-track tapes. If you'd like some 2-track suggestions let me know.

If you get any of these I'd be interested to hear your opinion of them.

Columbia:

The Dave Brubeck Quartet at Carnegie Hall, C2Q-587
Miles Davis Kind of Blue, CQ-379
Miles Davis Jack Johnson Soundtrack, SR-30455 (I have a spare copy of this tape that is missing several seconds of sound from start of side A/end of side B. If you would like this tape PM me your address and it's yours)
Thelonius Monk It's Monk's Time, CQ-644
Vivaldi, Four Concerti, Isaac Stern, David Oistrakh, Ormandy, Philadelphia Orchestra, MQ-404

Erato (Japan):

Vivaldi, Concertos Mandolines, Scimone, I Solesti Veneti, 7TS-7016RE

RCA Living Stereo:

Rimsky-Korsakoff, Scheherazade, Reiner, CSO, FTC2017

Everest:

Copland Appalachian Spring Ballet, Walter Susskind, London SO, T-43002

Verve:

Duke Ellington, Johnny Hodges, Side by Side, VSTC-237
Getz/Gilberto (with Jobim), VSTC-317

Impulse:

Count Basie and the Kansas City 7, ITC-304
Oliver Nelson/The Kennedy Dream, IPC-9144

Atlantic:

The Modern jazz Quartet, European Concert, ALP-1915
The Modern Jazz Quartet Plays George Gershwin Porgy and Bess, ALC-1935