Dynaco ST70 KIT - your opinion please
I am new to the hobby and in the process of trying out and learning about various amps. So far I have tried a few kits (gainclone and class D). Now it is tube time :-)
By coincidence I came across this Dynaco ST 70 kit:
Triode USA Dynaco ST-70 Tube DIY Amp Kit
I want to try it with various speakers, DIY Fostex FX120, DIY Jordan JX92s, DIY Wild Burro Betsy, commercial Vroemen Diva superiore. Frontend will be Bottlehead Quickie, 47 Treasure CD and vinyl.
What are your thoughts and recommendations about this? Any input is highly appreciated.
Michael's Project Page
I've got an original dynaco ST70 with the 7199 driver tubes. 1961 build with the fabric transformer insulation. It is a rugged beast with tens of thousands of hours on it, and decent sounding for 1961. I recently replaced the electrolytic caps (again, 3rd time) and the output tubes (third time) and it is back up to 35 W/ch and sounding original on my Peavey SP2-XT speakers. I've recently beaten the quallity of sound of the ST70 with 7199 drivers with a transistor amp modified for better cold bias current control, as does my road weary CS800s, so the 1% distortion Dynaco claimed for their circuit is about right. The triodeelectronics kit uses EF86 copy driver tubes, which might sound better than the unavailable 7199. The original design transformer had extra filament capacity to run two outboard two tube preamps, so the design has plenty of capability to run 4 driver tubes instead of two. The design has lots of aftermarket support, for example kevinkr has designed a 4 tube triode driver board sold by classicvalve.ca that he claims has a lot less distortion. My ST70 has been modified to derive the grid bias voltage from silicon rectifiers, not the original selenium one. I find the aftermarket high capacity cap boards a bit silly, as on classical music in my living room I'm running about 1.5 Vpp output max, so the semi-original power supply with 39 22 22 22 filter caps produces B+ just fine with no sags. This stays within the parameters of the 5AR4 rectifier tube, which I am still using due to the rumors about "cathode stripping" with silicon instant on B+ voltage. I've used the ST70 as a PA amp in a church occasionally, but didn't try to derive the high power out of it that you would want out of a bass guitar amp, for example. 35 W/ch works out to 17 Vac on 8 ohm, so the ST70 has plenty of power with fresh tubes and B+ cap for my 101 db@1W speakers. It was fine with my previous speakers, Peavey T-300 HF projectors, and KLH23 two ways. If you try to use acoustic suspension speakers like the old AR1's, the power out might be a little light for drum or cannon hits. I found over the years that if the ST70 sounded polite instead of real, the B+ voltage was sagging due to tired main caps, or the rectifier or output tubes were tired. I got about 8 years at 4 hr/night on the 5AR4, about 10 years on sets of 6CA7's, about 10 years per B+ cap.
One huge advantage of the tube design, mine has been struck by a lightning induced surge and only shorted across the power switch, and burned up the .01 uf turn off pop eliminator cap installed by me. No damage to the transformers. Try that with transistors without MOS supressors on the input. With the weather changing to be more extreme, you may soon have significant lightning storms in Europe.
I've bought parts form triodeelectronics.com, and find them helpful and nice to deal with. When they shipped me a mono volume pot instead of the stereo I ordered, they shipped the replacement to me at no charge.
So anyway, have fun.
Do you have any goals in mind as to what sound you're after? You can't go wrong with a Dynaco ST70 or its variants as they are good amps. Personally, I prefer the EL84-based PP amps, and single-ended triode amps based on 45, 2A3, or 300B. They are very transparent and musically involving. Just giving you some more ideas.
Strap the outputs for triode.
Add more capacitance to the second main B+ filter, like 4x or more.
Reduce the cathode resistors by a factor of 10x. (you need a DVM to set the bias, it will be 10x lower than the stock value, put a sticker on the chassis to show that. This will improve the bass.
You might consider buying the parts and getting an "aftermarket" driver board that uses something OTHER than the 7199 or the Russian equivalent for the input stage. That is to say the kit without the driver board...
These are the basic things that can be done to improve the basic amp.
I'd go for the Teflon wire option, IF you can figure out how to strip it without nicking the wire and/or cutting it accidentally (unless they are precut and stripped, doubt that).
This kit sounds like fun for more experienced people who like to try many variations. A good learning tool for a newbie. I use an EF86 in a guitar amp I built, and the measured gain is about 240, which is way more than I needed. In this amp you would need to either attenuate it's output or use rather large amounts of negative feedback. More feedback isn't necessarily better sounding in tube amps.
I researched, designed and built a similar amp using 6SN7's for the front end and phase splitter, (EL34 outputs) and use only about 12dB of negative feedback, and it sounds great. The hard core tube guys I know all seem to think low or no feedback, and triode mode output stage is the best sounding. I tend to agree. Triode mode output means lower source impedance driving the speaker, so tighter control of the speaker, and less un-calibration of a passive crossover, when feedback in minimal. More feedback makes the issue of phase margin harder to predict (tendency to oscillate or ring). Some people say high feedback makes an amp sound "constipated"... Check out my amp (the "musicbox") if interested at
My front end tube is the earlier equivalent of the 6SN7 (I forget the number). I thought the older tube shape with the grid cap was cool looking, so used it instead. Although not shown on my website, I since replaced that front end tube with a regular 6SN7 to reduce hum, and replaced the plate resostor with a constant current source for better power supply hum rejection, better linearity, and more gain. Don't bother with silver wire or Teflon insulation wire. Speaker wire that costs more than $50 is a joke too. Use 14 or 16 AWG wire for the speakers. It's the connectors that could be a weak link. I use gold alloy plated banana connectors (gold doesn't oxidize over time), and I solder the wires to these connectors rather than using the set screw that comes loose over time.
May I disagree?
Teflon insulated wire is a good idea for the inside wiring of the amp, IF you can properly strip it. There are a few methods that work.
For one thing, the simplest point, it doesn't melt and stink when you solder it.
Teflon insulated wire is great for newbies, keeps you from burning the insulation off when you move your eyes to the print or something and wave the soldering iron around. I'm not that new, but used a some 600V electrical tape in my H100 organ recap to repair damage to PVC insulation. I'm buying bulk teflon insulated wire from a suplus house, much cheaper than new silver plated stuff. Not legal in Euorpe, though.
I find the bass of my stock unit entirely adequate, actually. There are a million variations, the one I like the most adds two pots and splits the bias adjustment for the push-pull tubes and allows you to set them balanced again after they wear some, unevenly. Another thing I have done is run the B+ to a blank pin of the preamp power plug so I can monitor B+ while it is playing without taking it apart and tipping it on its back. That is now I know I don't have a B+ problem now at the listening levels I am using with 39-22-22-22 caps. Even when the cannon goes off in 1812 overture.
ok, a different driver circut, non pentode.
afaik almost all teflon insulated wire is silver plated copper.
there is some botique wire made now that is copper with teflon insulation.
Some comments, If I may. The Triode Electronics driver board gets populated with 2X EF86s and a single twin triode. That twin triode may be an ECC99, 12BH7, or 12AU7. I advise installing an ECC99, for its high gm. High gm in designs employing GNFB, like the ST70, provides resistance against slew limiting associated with the HF error correction signal.
As the ST70 uses "fixed" bias and an instant on, SS rectified, C- supply, even the slightest delay in B+ rise is sufficient to protect the O/P tubes against cathode stripping. A NTC inrush current limiting thermistor provides the short delay needed, if either SS diodes or a directly heated vacuum rectifier are/is employed in the B+ circuitry.
IMO, it's perfectly acceptable to SS rectify ST70 B+. You turn the excessively "tall" B+ rail from a liability into an asset, by employing combination bias. "Stand" the cathodes of each O/P tube pair on a 100 Ω/470 μF. network. A very convenient "idle" current test point is available and the original single bias adjustment for each channel is quite satisfactory. Matching requirements are reasonably relaxed. As is always the case, closeness in gm is a must for each pair. However, small differences in cathode current are acceptable.
An additional benefit of SS rectifying the B+ is the cooler running power trafo obtained by holding the value of the 1st filter cap. down to say 10 μF. Replace the overworked C354 choke with a Hammond 159T.
Hi Eli -- I'm interested in adding those thermistors to my mutant amp in an ST70 chassis. I'm using SS diode rectification, getting about +460V. What specs to look for in the NTC thermistors? What resistance spec? Small in value, like 100R or less? Or more like 1k? Thanks for the idea.
Oh, and where do they go in the circuit for slow ramp up of B+? Perhaps in the center tap of the power transformer?
|All times are GMT. The time now is 01:27 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio