2 VRMS digital sources are today's norm. That sort of signal will drive most designs into clipping and some attenuation is required.
Of course, you do need a preamp to play "vinyl". 45 or so dB. of net gain is about right for a 4 mV. MM level cartridge. Quite a few good designs for phono preamps can be found here at DIY Audio. FWIW, a tweaked version of RCA's passive EQ setup is what I'm associated with.
Attachments
Bryan
Eli made a wonderful suggestion in the Cit 2 and cit 5. I'm not sure of your budget but if you can't find a cit 2 or Cit 5, The Bon Latino amps are an option.
The kits that are reasonable and use high quality parts and are an excellent option and even tho they based on Dynaco architecture somewhat, they are far and away a big improvement as far as iron and circuit and board goes. Eli has additional tricks on those as well. Try for the Cit 2 or Cit 5 but if you don't feel comfortable, the Latino kit is very easy to build.
http://bob01605.50webs.com/
Eli made a wonderful suggestion in the Cit 2 and cit 5. I'm not sure of your budget but if you can't find a cit 2 or Cit 5, The Bon Latino amps are an option.
The kits that are reasonable and use high quality parts and are an excellent option and even tho they based on Dynaco architecture somewhat, they are far and away a big improvement as far as iron and circuit and board goes. Eli has additional tricks on those as well. Try for the Cit 2 or Cit 5 but if you don't feel comfortable, the Latino kit is very easy to build.
http://bob01605.50webs.com/
Last edited:
Thanks all for the suggestions and guidance.
While I'd love to find an old amp for restoration, I also like the "build", I'm a tinkerer... I build a LOT of my gear in life (Solar Roof Install, Brewery, Reverse Osmosis for Maple Sap concentration, Chicken Coop, Garden Shed, guitar amps, etc) and I find a lot of satisfaction enjoying the things that I build. So for me, this build falls into that category.
The process for me is the journey, getting to the destination, typing on forums, sourcing parts, posting results, tweaking. I simply love the journey.
I'll keep an eye out for an old Citation II or V for future reference. Where are good places to look?
Back to this project, I'm wondering if building on a breadboard will be beneficial. I'm still torn on the mono vs single chassis option... I'm mostly pro Monos other than the need for dual attenuation of volume.
I'm still leaning towards:
6922>ECC99>KT88 UL PP with regulated voltage on the input and G2 B+. SS rectification. EDCOR OTs.
What do you all think about the PS as provided? Would it benefit from inrush limiting or soft-start? How about a delay for the heaters to warm? Any other thoughts?
I'll continue to take notes, read up, devise questions and learn more as you all guide me.
Thanks!
Bryan
While I'd love to find an old amp for restoration, I also like the "build", I'm a tinkerer... I build a LOT of my gear in life (Solar Roof Install, Brewery, Reverse Osmosis for Maple Sap concentration, Chicken Coop, Garden Shed, guitar amps, etc) and I find a lot of satisfaction enjoying the things that I build. So for me, this build falls into that category.
The process for me is the journey, getting to the destination, typing on forums, sourcing parts, posting results, tweaking. I simply love the journey.
I'll keep an eye out for an old Citation II or V for future reference. Where are good places to look?
Back to this project, I'm wondering if building on a breadboard will be beneficial. I'm still torn on the mono vs single chassis option... I'm mostly pro Monos other than the need for dual attenuation of volume.
I'm still leaning towards:
6922>ECC99>KT88 UL PP with regulated voltage on the input and G2 B+. SS rectification. EDCOR OTs.
What do you all think about the PS as provided? Would it benefit from inrush limiting or soft-start? How about a delay for the heaters to warm? Any other thoughts?
I'll continue to take notes, read up, devise questions and learn more as you all guide me.
Thanks!
Bryan
I think I would separate pots to bias your KT88's mainly because you can get the correct bias current per device and your -54v may not be exactly -54v especially if you use you own transformer.
A single bias trim pot. per channel is OK, when a shared combination bias RC network is present. Some mechanism must be present to deal with the inevitable variances in an O/P tube pair.
Reputable vendors do a good job of matching PP pairs, but the tubes are not "identical". Even perfectly matched pairs rate to drift apart, as aging occurs.
Sorry what I am saying is that this amp does not have and adjustment for the the actual output current taken by both tubes. You can adjust the balance. There may be quite a spread on this even with KT88's or if you put in something else. You are also assuming you have -54v not -50v for example.
You could use one of the auto-bias boards. These adjust the bias automatically to the correct value. You can also use 100k with 6550 in this mode as it counts as self-bias. Just suggesting ideas to improve your amp. I've used the auto bias boards.
Last edited:
You don't regulate O/P tube g2 B+, when the UL taps on Edcor O/P trafos are employed. Combining regulated g2 B+ with UL mode is elegant and technically superior, but it requires hideously expensive, custom, O/P "iron" that contains separate, tertiary, screen grid windings.
A SS rectified bias supply turns on "instantaneously". Placing an inrush current limiting thermistor between rectifier and B+ PSU filter slows B+ rise down for a short interval that actually is adequate. The bias supply is already electrostatically shielding the warming O/P tube cathodes. 😉
Highly satisfactory results can be obtained from any of the several available rectification topologies. Implementation details rule the roost, regardless of which topology is settled on.
Monoblocks are fine. The "mantra" is no line stage gain in combination with 2 VRMS digital signal sources. Mating monoblocks to a buffered control center allows for close placement to the speakers and short speaker cable runs. Competent low O/P impedance/high current buffers drive long interconnect cabling without incident.
BTW, a passive control center can be fine. However, short, low capacitance cabling between center and power amplification is essential. No drive capability is present in a passive setup. Low capacitance means unshielded and braided, which is not the sort of thing to use in the cesspits of EMI/RFI associated with high density urban environments.
Thanks Eli for the clarification on UL and regulation. I see now that the suggestions was regulated G2 B+ with pentode mode, or UL without.
Sonically, what would the difference be between the two topologies assuming both were executed appropriately?
Sonically, what would the difference be between the two topologies assuming both were executed appropriately?
I completely agree with making "idle" current adjustable. Either an individual trim pot. for each tube or the combination bias "trick" that allows a single pot. in each channel to work well will get the job done. Reverse the schematic's diode polarity in this negative voltage situation.
Make the bias supply more than "tall" enough to completely cut the O/P tubes off. "Dial" the bias applied back, until the desired operating conditions are obtained.
Speaking of the negative bias rail, the provided 1/2 wave parallel multiplier works well in combination with a "spare" filament winding. More than 1 voltage can be taken from the multiplier and a small negative potential under 10M45S LTP tail loads is insurance against "compliance" trouble.
Attachments
Last edited:
Sonically, what would the difference be between the two topologies assuming both were executed appropriately?
Mr Duttman may be able to help you better on this.
The UL connection is designed to give the lowest distortion be providing local negative feedback to the screens. It also reduces the damping factor and should produce the most 'blameless' amp. However the screen voltage can be the limiting factor on the HT voltage and hence output power you get.
The pentode connection I know less about.
Mr Duttman may be able to help you better on this.
The UL connection is designed to give the lowest distortion be providing local negative feedback to the screens. It also reduces the damping factor and should produce the most 'blameless' amp. However the screen voltage can be the limiting factor on the HT voltage and hence output power you get.
The pentode connection I know less about.
Whatever the sonic differences are, listening to the specific device is needed. Given loop NFB's "homogenizing" effect, I think differences will be minimal. In order to obtain the same damping factor as UL yields with full pentode, more loop NFB is needed. Until damping factor matches, any comparison is bogus.
Triode mode is clearly different than the multi-grid modes and many regard triode as superior. 🙂 However, triode mode carries a 50+ % max. power penalty. 🙁 TANSTAAFL is forever with us.
Triode will have to wait for the day I finish the 14 year old 845 project...
Seeing as I’m looking for power to feed the ProACs, what would the expected output wattage be for UL mode versus pentode?
If I “need” the extra power, seems maybe I’m best with pentode.
Seeing as I’m looking for power to feed the ProACs, what would the expected output wattage be for UL mode versus pentode?
If I “need” the extra power, seems maybe I’m best with pentode.
I think UL gives you about 70-80% of the power of pentode for the same HT and screen voltage. For pentode power will drop as screen voltage is decreased. Some output valves are designed to run at lower screen voltages.
Ultra-linear - Wikipedia
The pentode mode can also allow you to use a higher HT as your not limited by the max screen voltage.
The phase shifts in the OPT tend to limit you to about 17dB GFB which will tail off above 10KHz due to the dominant pole in the loop. The UL tap partly cancels some of the distortion the output valves create reducing distortion before you apply GFB.
The UL has a big reduction in damping factor, The pentode mode is basically a current source driving the output transformer, so the damping factor is controlled by the GFB, the GFB tails off at higher freqencies so the damping factor will go up. This causes the pentode mode to often sound brighter than UL as the output is working into the inductance of the tweeter.
Hope this helps.
Ultra-linear - Wikipedia
The pentode mode can also allow you to use a higher HT as your not limited by the max screen voltage.
The phase shifts in the OPT tend to limit you to about 17dB GFB which will tail off above 10KHz due to the dominant pole in the loop. The UL tap partly cancels some of the distortion the output valves create reducing distortion before you apply GFB.
The UL has a big reduction in damping factor, The pentode mode is basically a current source driving the output transformer, so the damping factor is controlled by the GFB, the GFB tails off at higher freqencies so the damping factor will go up. This causes the pentode mode to often sound brighter than UL as the output is working into the inductance of the tweeter.
Hope this helps.
Last edited:
Most setups with a GNFB loop have a phase compensation capacitor. It's C6 and C12 in the Cit. 5 schematic. Those caps. have to be tuned to the specific O/P "iron" specimens being employed. Drive the amp with a 2 KHz. square wave and examine the speaker terminal O/P with an o'scope. Tweak the value of the phase compensation cap. to get the best looking O/P. It's a compromise among tilt, ringing, and overshoot.
A hefty amount of open loop gain is produced by Cit. 5 style small signal circuitry. That gain is definitely needed to drive both the "finals" and the GNFB loop. A benefit may be obtained from inductive wirewound resistors as the LTP plate loads. The HF "peaking" introduced can partially compensate for the natural HF roll off of the O/P trafo and that compensation reduces the GNFB's HF error correction signal level. The technique falls into the giving with 1 hand and taking with the other category.
A hefty amount of open loop gain is produced by Cit. 5 style small signal circuitry. That gain is definitely needed to drive both the "finals" and the GNFB loop. A benefit may be obtained from inductive wirewound resistors as the LTP plate loads. The HF "peaking" introduced can partially compensate for the natural HF roll off of the O/P trafo and that compensation reduces the GNFB's HF error correction signal level. The technique falls into the giving with 1 hand and taking with the other category.
Yep and there's also C15+R8 and R7 in parallel with the Ra of the first stage which forms the dominate pole for the amp. Its what is called lead-lag compensation.
Since the Ra of a pentode is high it works out this is at 9KHz. So the GFB will be -3dB lower at this point than at low frequency.
Since the Ra of a pentode is high it works out this is at 9KHz. So the GFB will be -3dB lower at this point than at low frequency.
Last edited:
It's likely cheaper to buy some new quality Output Transformers than buy a vintage HK that may have bad ones. The circuit used on the Heathkit W6M may be worth looking at also as it's got it's followers and all the tubes are in production.
When push comes to shove, the W6M is a Williamson style setup. There's always the issue of O/P "iron" quality, when Williamson style topology is implemented. I'm not saying that Edcor stuff's quality is insufficient, but a substantial risk would definitely be present. The "idiot resistant" nature of Mullard style is (IMO) a better bet.
Attachments
Hello Gentlemen,
Sorry for the grand absence. With Covid lessening in CT I went back to work and seems we've been making up for lost time... That and the chicken coop I build for the new birds and my kids, well, time flies.
After a much need vacation next week I'm back to the grind and will also start reviewing this project and beginning securing parts for testing.
Hope I didn't loose all the expert help and guidance with my delay. I also hope you are all well.
Thanks!!
Bryan
Sorry for the grand absence. With Covid lessening in CT I went back to work and seems we've been making up for lost time... That and the chicken coop I build for the new birds and my kids, well, time flies.
After a much need vacation next week I'm back to the grind and will also start reviewing this project and beginning securing parts for testing.
Hope I didn't loose all the expert help and guidance with my delay. I also hope you are all well.
Thanks!!
Bryan