Spectral DMC-10 mods

DHL

New Member
Apr 21, 2015
16
4
0
#1

I recently purchased a Spectral DMC-10 based on some of the posts I found here, and the fact that circuit schematics were available for this particular model. The sonic performance of this pre-amp is well revered, and I wanted a Class A discrete design that was simple, relatively low cost, but still capable of producing high performance. I am a tweaker at heart, so I can rarely buy something without giving it a once over and making improvements where I can. Although the design is quite good, there are some aspects that can be improved upon to make it better. Here is the summary of what I found for improvements.

1) Power supply.
The power supply for the DMC-10 is comprised of two sections. An AC transformer with a single pair of output windings in a remote chassis, and the DC rectifiers and filter capacitors in the preamp chassis (along with the low level preamp phono circuits). This is a major compromise in the design. Firstly, they run AC power to the amplifier chassis instead of filtered DC, which introduces AC noise into the low level signals. Secondly, since the transformer has only two output windings, to produce two separate V+/V- power supplies (one for Left Channel, one for Right Channel), they can only half wave rectify, which results in significant AC ripple and noise in the DC supply. There is also no ground reference from the AC supply, so this may potentially cause a safety issue if the amplifier chassis is not earth grounded via another means. My solution was to build another power supply that used two independent torroidal transformers, and supply two independent DC supplies from the remote chassis. More on this supply below.

2) Replace/upgrade the onboard filter caps and bypassing caps

3) Upgrade the voltage regulators. My unit is an early Alpha build, so some of these mods may not apply to later versions. On my unit the negative voltage regulators were 7905 series instead of the 337 series used in later models (which are the true match for the 317 positive regulators), so I replaced the 7905 with proper 337 parts. I also upgraded the heat sinks to longer versions. Spectral used some 400 ohm resistors to bypass the regulators (not sure why), and I removed these because they can "feed forward" power supply noise downsteam to the regulated voltages.

4) Replace unshielded phono wiring with shielded, twisted pair wiring from the RCA input jacks. On Alpha version Spectral ran unshielded wire from the RCA jacks to the PC board. Later versions ran PC traces directly to the jacks, with short wire connections (as with the other inputs). This exposed wiring (along with the AC from the transformer) was a significant source of hum in my amp, especially at higher gains.

5) Replace the polypro coupling caps. In the Alpha version, Spectral used what appear to be Plessey white box polypro coupling caps ( 2 x 4.7 mfd/ch) at the output of the phono stage and at the output of the preamp. In later versions, they replaced the phono stage output caps with in house packaged units (marked with their logo), but kept the white box coupling caps at the output. My suggested mod is to place both sets with 20 mfd WIMA polypro film caps. Turns out the mounting pin spacing for these caps was an identical fit to the OEM versions.

6) Replace the volume Alps pot with Arn Rotecap's Gold Point precision attenuators. My preference here, as I had a spare Gold Point 8K attenuator sitting around, and I have always preferred these discrete precision attenuators to any potentiometer. The original Alps pot had some noise issues, probably due to its age, and the Goldpoint is completely silent.

7) Replace the electrolytic caps in the protection circuitry with film caps. The protection circuit in this amp looks for RF and DC voltages. The filter caps are critical in the operation of this circuit. The bad news is the protection circuit has a relay that SHORTS the output to ground if a fault is detected, and if activated, may actually destroy the amp by overheating the output MOSFETs and burn resistors. That is why you see some of these units with burned PC boards. I also changed the output resistors to 400 ohms from 100 ohms to reduce any potential damage if this protection circuit is activated.

8) Replace the op amps on the tape output buffers. Spectral originally spec'd TL072 opamps for the tape output buffers. These are really poor performing opamps (OK back in the early 80's I guess). I have upgraded my unit with modern OPA627 high speed opamps more befitting the current performance of the DMC-10. However, one must be cognizant of the PS voltages, as the 627 (or the TL072 for that matter) have a MAXIMUM rating of +/- 18 volts. My DMC-10 exceeded that so I had to reduce the supply rails to +/- 17.5 volts.

9) Replace the bias current pots with multi-turn trimmers for more precise control

10) Replace the gain adjustment pot in the output buffer amp with a precision fixed resistor. I don't like pots in feedback loops, as Spectral has done to vary output buffer gain.

Here are some photos of my unit prior to mods:
IMG_1182 copy.jpg IMG_1178.jpg IMG_1180.jpg IMG_1181 copy.jpg



Photos of the modded DMC-10:
IMG_1201.jpg
Power supply chassis.

The new supply is two complete supplies, each with its own torroidal transformer, separate ground, and 90,000 mfd of filter capacitance. The white box in the picture is a time delay relay to soften turn on transients.

See part 2 for more photos
 

ack

VIP/Donor & WBF Founding Member
May 6, 2010
5,269
36
48
Boston, MA
#2
Nice work! Instead of three threads you can simply respond to this thread and post more pictures, in each response
 

marty

Well-Known Member
Apr 20, 2010
1,167
43
48
Far Hills, NJ
#4
DHL,
This is a fantastic post. There are many reasons why. But first and foremost, it shows your beautiful, meticulous and thoughtful upgrades with exquisite workmanship that any audiophile who knows how to use a soldering gun would be proud to call their own. I owned a DMC10 (delta and gamma versions). It did some things well that even back then, and revealed qualities that are still true of all Spectral gear today (high bandwidth, low distortion, quietness, etc). The phono stage in particular was actually quite good if you could get the DC offset to zero with a little help from good voltmeter and small screwdriver. It would not surprise me that your updates make your unit a formidable piece of gear that might compare favorably to many more expensive modern units. I love it when a unit like yours, rescued from the scrap heap of time, can be updated which might allow it to take on far more expensive Goliaths and compare sonically. I hope that is indeed the case. But the audiophile mindset is what is to be admired most of all. Great job. I take my hat off to you, sir.
Marty
 
Likes: CKKeung

DHL

New Member
Apr 21, 2015
16
4
0
#5
DHL,
This is a fantastic post. There are many reasons why. But first and foremost, it shows your beautiful, meticulous and thoughtful upgrades with exquisite workmanship that any audiophile who knows how to use a soldering gun would be proud to call their own. I owned a DMC10 (delta and gamma versions). It did some things well that even back then, and revealed qualities that are still true of all Spectral gear today (high bandwidth, low distortion, quietness, etc). The phono stage in particular was actually quite good if you could get the DC offset to zero with a little help from good voltmeter and small screwdriver. It would not surprise me that your updates make your unit a formidable piece of gear that might compare favorably to many more expensive modern units. I love it when a unit like yours, rescued from the scrap heap of time, can be updated which might allow it to take on far more expensive Goliaths and compare sonically. I hope that is indeed the case. But the audiophile mindset is what is to be admired most of all. Great job. I take my hat off to you, sir.
Marty
Marty:
Thank you very much for your kind words. I'm glad you can appreciate the time and effort I have invested in this project. With so much emphasis being placed by today's designers on IC based phono preamps, one almost has to resurrect older, classic products if one wants a Class A discrete design. This is simply due to the sad fact that low noise, discrete transistors are no longer being produced, as the mass market has gone almost completely to integrated circuit op amps. After studying the schematics of the DMC-10, it was clear to me that this circuit topography was superior to many of the early 80's designs, including Pass Labs/Threshold designs like the Xono (which I considered building). You are correct about the DC performance being outstanding, although I would prefer Spectral would have used a DC servo controller in the phono stage. I still get a bit too much offset and drift in my phono stage (50 or so mV at the output), which is still incredible given that the phono has a DC gain of over 1000! This computes to an offset at the input stage of 50 micro volts. The buffer amp stage (which is almost identical to the phono stage) routinely produces offsets of under 5 mV, which makes direct DC coupling to a line stage or power amp possible. But, of course a servo controller would have increased complexity, and potentially impacted sound quality (Nelson Pass and John Curl will disagree with me on this).

Even though I did not mention sound quality in the original post, I have had a few days to listen to the pre-amp (mainly the phono), and I am extremely pleased with the sound. Very smooth and dynamic, with no harsh brittleness or sibilance on female close mic'd vocals or symbols. Phono hum levels have been considerably reduced at higher gain settings.
 
Likes: CKKeung

marty

Well-Known Member
Apr 20, 2010
1,167
43
48
Far Hills, NJ
#6
Marty:
Thank you very much for your kind words. I'm glad you can appreciate the time and effort I have invested in this project. With so much emphasis being placed by today's designers on IC based phono preamps, one almost has to resurrect older, classic products if one wants a Class A discrete design. This is simply due to the sad fact that low noise, discrete transistors are no longer being produced, as the mass market has gone almost completely to integrated circuit op amps. After studying the schematics of the DMC-10, it was clear to me that this circuit topography was superior to many of the early 80's designs, including Pass Labs/Threshold designs like the Xono (which I considered building). You are correct about the DC performance being outstanding, although I would prefer Spectral would have used a DC servo controller in the phono stage. I still get a bit too much offset and drift in my phono stage (50 or so mV at the output), which is still incredible given that the phono has a DC gain of over 1000! This computes to an offset at the input stage of 50 micro volts. The buffer amp stage (which is almost identical to the phono stage) routinely produces offsets of under 5 mV, which makes direct DC coupling to a line stage or power amp possible. But, of course a servo controller would have increased complexity, and potentially impacted sound quality (Nelson Pass and John Curl will disagree with me on this).

Even though I did not mention sound quality in the original post, I have had a few days to listen to the pre-amp (mainly the phono), and I am extremely pleased with the sound. Very smooth and dynamic, with no harsh brittleness or sibilance on female close mic'd vocals or symbols. Phono hum levels have been considerably reduced at higher gain settings.
Yes, I've had the conversation about servo controlled DC offset with Rick Fryer and you are correct, he believed (at the time) that any servo circuit for such an application would result in sonic compromise. I then complained about the fact that the DC offset does drift periodically (and with temperature of course, which is a small issue in a true 4 season environment) and it would be nice not to have to take out the voltmeter and screwdriver every once in a while to adjust the pots. That's when he admonished me and said that this should never be done by the user, and only the Spectral factory is knowledgable enough to know how to do that. That was my first encounter with the legendary arrogance of Rick Fryer. I was indeed tempted to tell him to consider adjusting his medications, but realized that would be non-productive. That said, he and Keith make a hell of a product and in fact I currently use Spectral RS400 monoblocks in my system (albeit with a VTL 7.5 III preamp). What I also admire about them is that they could easily charge more for the products they make, but they do not. It is well known that Fryer is independently wealthy, and you have to admire a manufacturer that really is not as much profit oriented as they are value oriented to the consumer.
 
Likes: CKKeung

DHL

New Member
Apr 21, 2015
16
4
0
#7
Yes, I've had the conversation about servo controlled DC offset with Rick Fryer and you are correct, he believed (at the time) that any servo circuit for such an application would result in sonic compromise. I then complained about the fact that the DC offset does drift periodically (and with temperature of course, which is a small issue in a true 4 season environment) and it would be nice not to have to take out the voltmeter and screwdriver every once in a while to adjust the pots. That's when he admonished me and said that this should never be done by the user, and only the Spectral factory is knowledgable enough to know how to do that. That was my first encounter with the legendary arrogance of Rick Fryer. I was indeed tempted to tell him to consider adjusting his medications, but realized that would be non-productive. That said, he and Keith make a hell of a product and in fact I currently use Spectral RS400 monoblocks in my system (albeit with a VTL 7.5 III preamp). What I also admire about them is that they could easily charge more for the products they make, but they do not. It is well known that Fryer is independently wealthy, and you have to admire a manufacturer that really is not as much profit oriented as they are value oriented to the consumer.
Unfortunately, many manufacturers have to assume that their customers fit the "lowest common denominator", and that if you allow them access to the inner workings, they will screw them up. I find this attitude unfortunate, but sort of understand why they present it. I myself, having a reasonably good understanding of audio circuitry, have been treated this way and resent it. I also do not care for the "secrecy" and the refusal of manufacturers to publish circuit schematics. It's not that I want to copy and reproduce their products as a competitor, but I want to know exactly what I am buying, and how I think the product will function. Or if I can (or need) to make improvements. I stopped buying PS Audio products for this reason. They misrepresent the circuit topography, and you will not know until you get the product, open it up, and find out its not all Class A discrete as they claim. If they want to protect their novel circuitry, they should patent it, and then they can publish the circuit diagram without fear of copy.

The servo controller controversy has been around for many years. I recall a post in the DIY Audio forums where Nelson Pass was arguing that use of the cheapest cheese ball op-amp was inaudible, and that he used the lowest grade part just so tweakers would be discouraged from upgrading it (the logic of this escapes me too). A careful engineering analysis (one that does not assume all op-amps are ideal devices) shows that the amplifier properties will indeed add some level of distortion, and the dilemma that ensues is just how much (and of what character) is objectionable. That is why I prefer to upgrade these circuits to the best known performing op-amps, just in case. What can it really hurt?

With respect to the offset of the phono stage, it's capacitor coupled anyway so unless the offset is a few hundred millivolts or more, it has little impact on a amp that can put out 10 volts plus. I was curious as to how low the offset could be set, because I considered putting in jumper terminals to bypass the output coupling caps in the phono stage, making the DMC-10 truly DC coupled input to output. The output buffer amp, that has a much lower overall DC gain, can easily be direct coupled because the offsets can be adjust down to the single digit millivolt levels and they do not drift. Unfortunately, I could never get the phono stage offset much below 50 mV, and with a buffer stage gain of about 10, direct coupling would produce an output offset of 500 mV, way too much to feed to another external amplifier. So I have left the coupling Wima 20 mfd caps in the circuit, and from what I can tell, these have little negative impact on the sound quality.

Another critical adjustment is the output stage bias current. The factory specs a nominal 20 mA, and you need to stick with this because you can destroy the output MOSFETs at a current much higher than 20 mA (due to secondary breakdown considerations). When I got my amp, it was running at 30 mA. My power supply voltages were also "off", with the negative rail over 18.5 V and the positive rail at 17.3V. After replacing the regulators (as I mentioned) and putting in the precisely correct voltage adjustment resistors, I got +/- 17.5 V.
 
Likes: CKKeung

Ron Resnick

Site Co-Owner, Administrator
Jan 25, 2015
5,381
163
63
Beverly Hills, CA
#9
Welcome to WBF, DHL!

Thank you for your fantastic post!
 
Jun 18, 2011
801
18
18
Hong Kong
#10
Hello DHL,
Your work on the DMC-10 is fantastic!
Please post and share more often on WBF.
Many thanks!
 
Jun 5, 2010
838
5
18
#11
Wow DHL! Your workmanship is impeccable. Your mods look original, as if the unit was untouched.

I modded my DMC 20 by replacing the cables from the input and outpit jacks to and from the circuit board to MIT hook up cables.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. A place where audiophiles and audio companies discuss existing and new audio products, music servers, music streamers and computer audio, digital to audio convertors (DACS), turntables, phono stages, cartridges, reel to reel, speakers, headphones, tube amplifiers and solid state amplification. Founded in 2010 What's Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals we enjoy learning about new things and meeting new people and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing