Rh84 built, first impressions

[Tubeglowpio]

So I got around to finishing the rh84 build, it was built as the circuit describes besides some power supply tweaks and plate voltage ended up around 276 volts. Cathode current ended up being around 34 ma so only dissipating about a little over 9 watts. I also added a 100k pot on the input and removed the 1meg bleeder resistor. The positive things are that the amp is dead quiet and it sounds clean all the way turned up. Downsides is the gain is pretty low when using my phone and just OK when using my computer. Haven't tried my turntable yet. I'm using 95db speakers btw. it gets about half as loud as my Maggie 8601 amp builds, maybe a tad over half. The biggest problem that I have is the bass is on the weak side and the mids are kind of hollow and not thick and juicy like I expected. Any thoughts or help to improve the tone of this thing?

 

[dowto1000]

Hello !!

Sorry to hear this. Need to put this right for your family member.

Post a schematic of YOURS, with voltages, so we all can see precisely what you have built and talk specifics.

What parts types, wire, did you use? In SE, all is critical.

How about below deck photos??

Nine Watts dissipation of a 12 Watt Maximum Plate is 75% of maximum, which, IF all else is built AOK, should not present any problem, sonically IMHO. ( I regularly get very good results at 62%, or 7.5 Watts dissipation, the Golden Ratio percentage . )

Dowto1000

 

[Tubeglowpio]

Hello !!

Sorry to hear this. Need to put this right for your family member.

Post a schematic of YOURS, with voltages, so we all can see precisely what you have built and talk specifics.

What parts types, wire, did you use? In SE, all is critical.

How about below deck photos??

Nine Watts dissipation of a 12 Watt Maximum Plate is 75% of maximum, which, IF all else is built AOK, should not present any problem, sonically IMHO. ( I regularly get very good results at 62%, or 7.5 Watts dissipation, the Golden Ratio percentage . )

Dowto1000

Hey Jeff, ive been really busy with my physics class and work this summer, finally had a few hours this morning to finish the thing and was a bit let down, maybe just not my flavor. I want to check voltages again because I forgot that I checked those with my dim bulb tester in line. Already rechecked the bias and now I've got 40ma running so probably closer to spec on everything now. Nothing is wrong with the amp, worked great right away, no audible distortion even all the way up and pretty smooth hi-fi sounding just missing a little bass and the mids are not as 3d and lush as I would have thought for a decent tube amp. Thing is also dead quiet even on my 95 db speakers. It's a bit on the rats nest side as far as wiring, not proud of it but it's because I had the amp laid out for a different amp design so worked around it. I listened a bit more and it's sounding a tiny bit better because I pretty sure the caps are breaking In, the Russian pio caps seem to take some hours to bloom and I also ended up using a pair of solens for the preamp power supply section, I told everyone I wouldn't but was what I had on hand. I heard they mellow out over time so there's that. I guess the only thing I really questioned was if wiring the 100k pot into the 220 grid resistor and dropping the 1 Mohm resistor could have caused any cut in gain or more likely I was expecting too much gain out of a 12at7 that just doesn't have the gain like a 12ax7 which I'm used to.

 

[dowto1000]

Is your supply choke a Hammond? That is what I use. What is the model number of your choke(s) , and or, it's HY and, most important of all, its measured DCR or DCR rating??

Solen film caps, IMHO, are inferior. Far far better , use WIMA's DC LINK MKP-4 caps with FOUR PINS ( not 2 ).

What resistor types are you using ?? Brand and wattage. All of this is critical in SE executions.

Wire size and quality greatly effects bass response in a SE amp. Most DIYers are unaware of how to optimize this, nor comprehend the need to do so. Wiring actually starts with the AC plug and power cord AWG ....and wire type !!

In a SE amp, if you want quality sound and good dynamics, each Ra and Rk needs to be configured as paralleled pairs of Rs, matched to EACH OTHER within 0.01%, so there are no possible skews.

There is a sonic advantage to this. Two Rs in a SE amp always will sound better than one, if MATCHED tightly ( so there are no skews ).

I read on line today, Jeff Day testing Lefson high quality Rs ( about 30, 60, 90 bucks, for each R ) and the difference between the three was, the " cheapest resistor " was a single element, and had single ( silver ) lead outs . The " middle priced Rs " had the same single exact resistive element, but had DOUBLED-UP ( silver ) lead outs, the only difference!! The highest quality resistor had TWO resistive elements, ( each at twice the desired ending value - placed in parallel ), and it also had doubled-up ( silver ) lead outs.

Go read what Jeff had to say, about what he heard !!

https://positive-feedback.com/audio...end-analog-audio-devices-the-lefson-resistor/


The take away, of Jeff's article is what I suggest, use two high quality Rs for each Ra and Rk, matched very tightly in a SE amp execution. I use paralleled and 0.01% matched Rs, sometimes up to three paralleled. ( See my 6AQ5 Topology schematic, posted yesterday, and my Avatar. )

Let me know what choke you are hearing !! YES, the amp needs to break in, caps, solder joints, etc. 100-200 hours will tell the tale much better.

The gain would be better if the driver stage's Rk was film cap bypassed, properly, but that costs lotsa $$$ to do it well, so the designer skipped it !!

Dowto1000

 

[Tubeglowpio]

Is your supply choke a Hammond? That is what I use. What is the model number of your choke(s) , and or, it's HY and, most important of all, its measured DCR or DCR rating??

Solen film caps, IMHO, are inferior. Far far better , use WIMA's DC LINK MKP-4 caps with FOUR PINS ( not 2 ).

What resistor types are you using ?? Brand and wattage. All of this is critical in SE executions.

Wire size and quality greatly effects bass response in a SE amp. Most DIYers are unaware of how to optimize this, or comprehend the need to do so. Wiring actually starts with the AC plug and power cord AWG ....and wire type !!

In a SE amp, if you want quality sound and good dynamics, each Ra and Rk needs to be configured as paralleled pairs of Rs, matched to EACH OTHER within 0.01%, so there are no possible skews.

There is a sonic advantage to this. Two Rs in a SE amp always will sound better than one, if MATCHED tightly ( so there are no skews ).

I read on line today, Jeff Day testing Lefson high quality Rs ( about 30, 60, 90 bucks, for each R ) and the difference between the three was, the " cheapest resistor" was a single element, and had single ( silver ) lead outs . The " middle priced Rs " had the same single exact resistive element, but had DOUBLED-UP ( silver ) lead outs, the only difference!! The highest quality resistor had TWO resistive elements, ( each at twice the desired ending value - placed in parallel ), and it also had doubled-up ( silver ) lead outs.

Go read what Jeff had to say, about what he heard !!

https://positive-feedback.com/audio...end-analog-audio-devices-the-lefson-resistor/


The take away, of Jeff's article is what I suggest, use two high quality Rs for each Ra and Rk, matched very tightly in a SE amp execution. I use paralleled and 0.01% matched Rs, sometimes up to three paralleled. ( See my 6AQ5 Topology schematic, posted yesterday. )

Let me know what choke you are hearing !! YES, the amp needs to break in, caps, solder joints, etc. 100-200 hours will tell the tale much better.

Dowto1000

Both are Hammond chokes, they were the ones I modeled in psud2. Most of the resistors are kiwame, a few ohmite carbon comps were used for the grids. I did see that your working on a 6AQ5, also need to see how that 45 is going. Like I said just been too busy, wish I had more free time to work on that 2A3 amp right now but will have some time in about 5 weeks. I have experienced break in time with my other amps just expected more from this amp already. I could see people enjoying it, sounded a little solid state hifi to me though so far. I haven't heard too many tube amp yet but my Maggie has some of that magic midrange that I like, we'll see how this amp gets over the next week. I'm planning on giving it to my dad on the 4th of July so will break it in while I build a pair a speakers tailored to this amp.

 

[dowto1000]

We do not know what you modeled in PSUD What model number Ls are they? I can look up their specs . What is the precise configuration to the Finals, of Ls and Cs ???

What is the measured or simulated mVAC ripple, at the C, feeding the Finals tube?

I won't build Type 45 SETs anymore.

When well executed, and mass - loaded with about 60 pounds of concrete on the chassis' top, I can not stand listening to piano with any existing vintage Type 45 ( not an EML ) . When a vintage 45 tube is totally amp optimized, it has a broad and very shallow suck out, centered in the midrange, which causes me to give up all-hope for a Type 45.

A JJ 2A3 - 40, simply "smokes" any vintage 45 in my Directly Coupled SET amp executions, on the ALTEC VOTT A7-8s.

Jake, I believe the 6AQ5 UL SE amp I intend to build can possibly out-perform ALL vintage 45 tube amps ever built. I shall hopefully - soon see !! A $5.00 tube. What a hoot.

Thanks.

Dowto1000

 

[dowto1000]

Jake,

You state " the mids are kind of hollow and not thick and juicy like I expected. Any thoughts or help to improve the tone of this thing? "

Well, after sleeping on this overnight, I figured out Sunday morning, it came to me, what you are hearing. You are hearing the Cs in your supply, those Solen film caps, and their deficiencies at playing wide band linearly, with proper power..

NO single cap will produce a linear frequency and power response.

IME, all caps need help, by carefully bypassing the single cap with well-chosen added film bypasses, smaller uF values.

Some of this information is proprietary, so I will send you a PM sometimes today, as to what to add and listen to, in parallel with the Solen supply caps. You will LEARN much from what you do and hear. Its not fully a matter of parts break-in time Jake, but rather, proper amplifier design !!

In the end, we all listen to a modulated power supply. How good is it ??


Dowto1000

 

[Tubeglowpio]

Jake,

You state " the mids are kind of hollow and not thick and juicy like I expected. Any thoughts or help to improve the tone of this thing? "

Well, after sleeping on this overnight, I figured out Sunday morning, it came to me, what you are hearing. You are hearing the Cs in your supply, those Solen film caps, and their deficiencies at playing wide band linearly, with proper power..

NO single cap will produce a linear frequency and power response.

IME, all caps need help, by carefully bypassing the single cap with well-chosen added film bypasses, smaller uF values.

Some of this information is proprietary, so I will send you a PM sometimes today, as to what to add and listen to, in parallel with the Solen supply caps. You will LEARN much from what you do and hear. Its not fully a matter of parts break-in time Jake, but rather, proper amplifier design !!

In the end, we all listen to a modulated power supply. How good is it ??


Dowto1000

Yeah, that could be part of it, I did bypass the second filter cap but with another solen haha. Im going to try bypassing with a pio and see if there is any difference. I do think they will mellow out on their own a bit as a couple hours in and it's starting to at least sound listenable. But yeah feel free to pm, don't have much time today but will try to hop on here later today at some time.

 

[primosounds]

Hey Jeff, ive been really busy with my physics class and work this summer, finally had a few hours this morning to finish the thing and was a bit let down, maybe just not my flavor. I want to check voltages again because I forgot that I checked those with my dim bulb tester in line. Already rechecked the bias and now I've got 40ma running so probably closer to spec on everything now. Nothing is wrong with the amp, worked great right away, no audible distortion even all the way up and pretty smooth hi-fi sounding just missing a little bass and the mids are not as 3d and lush as I would have thought for a decent tube amp. Thing is also dead quiet even on my 95 db speakers. It's a bit on the rats nest side as far as wiring, not proud of it but it's because I had the amp laid out for a different amp design so worked around it. I listened a bit more and it's sounding a tiny bit better because I pretty sure the caps are breaking In, the Russian pio caps seem to take some hours to bloom and I also ended up using a pair of solens for the preamp power supply section, I told everyone I wouldn't but was what I had on hand. I heard they mellow out over time so there's that. I guess the only thing I really questioned was if wiring the 100k pot into the 220 grid resistor and dropping the 1 Mohm resistor could have caused any cut in gain or more likely I was expecting too much gain out of a 12at7 that just doesn't have the gain like a 12ax7 which I'm used to.

Hi TGP. i have made over 10 of these amps and i have not experienced the problem you have especialy with bass. I think you should remove the pot and put back the 1meg grid to ground resistor. As you already suspect you have lost some input signal with the grid being at less than 100k. Actually, you can just put the 1M back and leave the pot .
I use Solens in the PS all the time and there is a break in period but it is mostly in terms of a type of edge or hardness to the sound rather than a frequency issue.
Did you use the LM317 for biasing the output tubes? Doing that gives better bass dynamics than the cathode resistor.

 

[dowto1000]

Hi TGP. i have made over 10 of these amps and i have not experienced the problem you have especialy with bass. I think you should remove the pot and put back the 1meg grid to ground resistor. As you already suspect you have lost some input signal with the grid being at less than 100k. Actually, you can just put the 1M back and leave the pot .
I use Solens in the PS all the time and there is a break in period but it is mostly in terms of a type of edge or hardness to the sound rather than a frequency issue.
Did you use the LM317 for biasing the output tubes? Doing that gives better bass dynamics than the cathode resistor.

Hi Primo,

I fully agree with you, in that I would like to see maybe 400K to 500K , as the input tube's grid-to-ground value, and there would be no harm, would be instructive to Jake, in trying the 1 meg, in series with the pot. The 100K pot somewhat shorts-out the input signal, before it even gets to the grid of the input tube.

Dowto1000

 

[Tubeglowpio]

Hi TGP. i have made over 10 of these amps and i have not experienced the problem you have especialy with bass. I think you should remove the pot and put back the 1meg grid to ground resistor. As you already suspect you have lost some input signal with the grid being at less than 100k. Actually, you can just put the 1M back and leave the pot .
I use Solens in the PS all the time and there is a break in period but it is mostly in terms of a type of edge or hardness to the sound rather than a frequency issue.
Did you use the LM317 for biasing the output tubes? Doing that gives better bass dynamics than the cathode resistor.

Hey primo, so I may have jumped the gun a bit. I honestly posted that about an hour after first test and was pretty tired so was caught off guard by my expectations, I'm used more bass in my other builds and bass is there in this one but just not as much and since the treble has a slight boost it makes it also seem like there is less bass. I think since the Maggie got me into tubes it's been my standard as far as bass and thick 3d mids, that's what the Maggie did well in my system, it was not a very balanced amp though till I modded it. I don't see how the 1 meg will help if I have the pot in there because the pots resistance is ten times lower so it will dominate the input right? The two impedances would be in parallel so it would still end up close to 100k right? I've used a 100k pot on the Maggie but also had it go through a coupling cap then had a 1 meg bleeder resistor and the bass was fine.

I just used a cathode bias because I had already installed pots to adjust the bias. Did you end up bypassing the solens? I do think it's the hardness that bothers me, especially on my Klipsch since those are forward sounding anyways. I also have a switch for ultra linear and it sounded slightly smoother btw.

 

[Tubeglowpio]

Hi Primo,

I fully agree with you, in that I would like to see maybe 400K to 500K , as the input tube's grid-to-ground value, and there would be no harm, would be instructive to Jake, in trying the 1 meg, in series with the pot. The 100K pot somewhat shorts-out the input signal, before it even gets to the grid of the input tube.

Dowto1000

when you say in series with the pot do you mean grid of tube to ground or do you actually mean in series for the pots wiper to the grid?

 

[Tubeglowpio]

Hi Primo,

I fully agree with you, in that I would like to see maybe 400K to 500K , as the input tube's grid-to-ground value, and there would be no harm, would be instructive to Jake, in trying the 1 meg, in series with the pot. The 100K pot somewhat shorts-out the input signal, before it even gets to the grid of the input tube.

Dowto10

Hi Primo,

I fully agree with you, in that I would like to see maybe 400K to 500K , as the input tube's grid-to-ground value, and there would be no harm, would be instructive to Jake, in trying the 1 meg, in series with the pot. The 100K pot somewhat shorts-out the input signal, before it even gets to the grid of the input tube.

Dowto1000

I think part of it is just that the 12at7 doesn't have much gain so this amp would be better driven by a preamp, I am impressed on how quiet this build turned out as far as noise, I cant hear anything even with my ear touching the driver and the sound is really clean and pretty smooth besides the sharpness to the vocals and that. Soundstage is the worst of all my amps though so far, it fills the room ok but I cant find that sweet spot where the voices sound smack dab floating in the middle. I'm going to let it break in a bit more then I'll put it on the scope and see if I notice any roll of on the bottom end and any peaks.

 

[primosounds]

Replacing the 1M resistor with a 100k pot means even at full open there is only 100k of grid to ground resistance. So try an experiment and add 500k or 1m resistance in the ground leg of the pot. You dont have to solder it in since you are just experimenting. In the RH84 write up Alex Kitic, does mention that the amp does work best with a preamp as the input sensitivity is right at 1v for full power, which is on the lower side of things.
Regarding your lack of soundstage you might check the speaker phasing just to be sure as my experience is that this amp is excellent in that regard. Also bypassing the Solens with a 1.0uf x 400 or 600vdc PIO cap will definitely smooth things out. I have done that before and after a few months i removed the bypass and didn't notice any difference attributing that to the Solen's finally loosening up.
What kind of at7 are you using. I actually prefer the 5965 which seems to sound a little better.
I did try a few different ones before settling on one. There are many other at7 types that are worth tube rolling.
Finally, what is the cathode bias resistor and cap combo that you are using. As Jeff mentioned, every part in a SE amp is critical so keeping the same quality components at the cathode is important. I also always bypass the electrolytic cap with a film cap and using a very HQ bypass electrolytic cap. A Mills resistor is an excellent choice and i have also tried Aircol and Dale with good results.

 

[dowto1000]

Replacing the 1M resistor with a 100k pot means even at full open there is only 100k of grid to ground resistance. So try an experiment and add 500k or 1m resistance in the ground leg of the pot. You dont have to solder it in since you are just experimenting. In the RH84 write up Alex Kitic, does mention that the amp does work best with a preamp as the input sensitivity is right at 1v for full power, which is on the lower side of things.
Regarding your lack of soundstage you might check the speaker phasing just to be sure as my experience is that this amp is excellent in that regard. Also bypassing the Solens with a 1.0uf x 400 or 600vdc PIO cap will definitely smooth things out. I have done that before and after a few months i removed the bypass and didn't notice any difference attributing that to the Solen's finally loosening up.
What kind of at7 are you using. I actually prefer the 5965 which seems to sound a little better.
I did try a few different ones before settling on one. There are many other at7 types that are worth tube rolling.

Jake,

Primo and I are suggesting the same thing essentially.

A 100K total possible grid to ground value on the RCA input, due to the pot value choice, is too low !!!! It might be a bit better with 500K pot. 100K Ohms shorts out the signal before it gets to the grid of the driver. As suggested, add 500K or 1 M to the ground return of that pot, likely 500 K . That will do fine.

If you ( or your Dad )have a volume control on your sources, by all means, I whole heartedly suggest to get RID of ANY pot on the intake of the amp. Its a terrible novice idea. It ruins fidelity, right from the get go. Why ?? Due to wiper-resistive element rectification, and, all the extra unneeded garbage in the circuit. This is the LOWEST level of signal in the entire amp. KISS rules in SS design.. With no pot, I 'd use a 400K - 500K quality ( not CC ) R, Alex uses 1 M.

Get the driver tube freely, lively, to swing, it needs a high impedance circuit there.

Dowto1000

 

[primosounds]

I think Alex chose 1M to preserve as much signal as possible for the input grid. Since the design has somewhat "low" sensitivity. This is not an issue if using a powered preamp as the output of the preamp will easily supply more than 1v of input signal.

 

[dowto1000]

Yeah, that could be part of it, I did bypass the second filter cap but with another solen haha. Im going to try bypassing with a pio and see if there is any difference. I do think they will mellow out on their own a bit as a couple hours in and it's starting to at least sound listenable. But yeah feel free to pm, don't have much time today but will try to hop on here later today at some time.

TRY a 0.22uF PIO for the mids. No PIO has worked well enough for me, in recent years. May become too syrupy. Don't know. That value WILL effect the mids Jake !

Dowto1000

 

[dowto1000]

I think Alex chose 1M to preserve as much signal as possible for the input grid. Since the design has somewhat "low" sensitivity. This is not an issue if using a powered preamp as the output of the preamp will easily supply more than 1v of input signal.

Good on your first sentence !!! A++ !! . The amp will sound better, despite a preamp's possible greater signal level, if the first stage has a high grid to ground. This is NOT just a function of the overall preamp - amp -system gain. We want the first stage to " swing" - be lively, naturally. 100K is not going to allow that.

Dowto1000

 

[Tubeglowpio]

Replacing the 1M resistor with a 100k pot means even at full open there is only 100k of grid to ground resistance. So try an experiment and add 500k or 1m resistance in the ground leg of the pot. You dont have to solder it in since you are just experimenting. In the RH84 write up Alex Kitic, does mention that the amp does work best with a preamp as the input sensitivity is right at 1v for full power, which is on the lower side of things.
Regarding your lack of soundstage you might check the speaker phasing just to be sure as my experience is that this amp is excellent in that regard. Also bypassing the Solens with a 1.0uf x 400 or 600vdc PIO cap will definitely smooth things out. I have done that before and after a few months i removed the bypass and didn't notice any difference attributing that to the Solen's finally loosening up.
What kind of at7 are you using. I actually prefer the 5965 which seems to sound a little better.
I did try a few different ones before settling on one. There are many other at7 types that are worth tube rolling.
Finally, what is the cathode bias resistor and cap combo that you are using. As Jeff mentioned, every part in a SE amp is critical so keeping the same quality components at the cathode is important. I also always bypass the electrolytic cap with a film cap and using a very HQ bypass electrolytic cap. A Mills resistor is an excellent choice and i have also tried Aircol and Dale with good results.

Well I have the ground sides of the pot tied together then to the drivers star ground so I would have to desolder the ground and insert the 470k resistor there? or should I solder a 470k resistor from each of the 12at7's grids to ground? I'm using a Mullard 12at7 so would think that would help mellow it out a bit. I think I may have came off a little harsh about the amp, It has good qualities, sounds pretty hifi just too bright in my system right now but I'm going to mess with it after I burn it in a bit more and I will try what you guys suggest, feel free to draw a diagram, I'm a little bit confused on where you want me to try that resistor now, ground leg to ground with existing ground desoldered for test or just clip a 470k to each grid and ground? Thanks guys, hope I'm not bashing the RH84 too much, I think it's mostly cap choice, speaker choice, input impedance design choice, burn in and my personal tastes in sound combined on this one so far.

 

[Tubeglowpio]

Good on your first sentence !!! A++ !! . The amp will sound better, despite a preamp's possible greater signal level, if the first stage has a high grid to ground. This is NOT just a function of the overall preamp - amp -system gain. We want the first stage to " swing" - be lively, naturally. 100K is not going to allow that.

Dowto1000

Thanks, I honestly didn't know that resistor had as much impact on the sound, I knew it helped with gain and played a role in impedance but I'm learning something new.