The phase splitter currently activating my attention is the isodyne by E.F Worthen. There seems alot to commend it, although the article suggests matched tubes, the real drawback is swapping tubes can upset the running. I've never been able to get this configuration to run without having to re-tweak the circuit values. This is where the basic concertina over-scores.
Anyone else had a crack with the Worthen design ?
richy
Anyone else had a crack with the Worthen design ?
richy
Only in simulation, out of curiosity. It's basically a parrotface splitter with an attempt to make it reasonably linear, but I agree with slideman82 - it's got too many tubes. It's the opposite of the KISS principle, IMHO, i.e unnecessarily complicated for successful phase splitting (or inversion, if you prefer).
The conclusion I've reasched with splitters/inverters is that if I can't achieve what I want with either the cathodyne/concertine/split-load or the LTP/catode-coupled/Schmitt/differentail splitter designs, then I'm wasting my time.
The combination of gain stage feeding a cathodyne will give you more 2nd harmonic than an LTP.slideman82 said:
In other words, the gain stage provides the 2nd harmonic, and the cathodyne adds practically nothing to that. An LTP inherently produces 3rd harmonic.
However, a paraphase will probably give you the most non-linear distortion, including 2nd harmoinc, and a rich sound.
I took a look at the Vox AC50 schematic, and realised it had a floating paraphase phase splitter (http://www.freewebs.com/valvewizard/floatingparaphase.html). I think this type is not pretty linear, and I know I've seen it in old guitar amps like first Fender Twins.
Can anyone give an opinion about this phase splitter?
Can anyone give an opinion about this phase splitter?
These types of phasesplitters (aka ECC83 types) don't like to be output loaded; as thd rockets with rising frequency, so following this a low u Williamson diff driver is often used, coupling to the finals. However, modifying a standard paraphase with separate cathode resistors, global nfb can be injected into the first stage cathode and the GEC 88-50Watt HiFi amp has exactly this all put straight to provide 0.2% thd at 50W o/p at around 20dB global nfb. So we have a 3 stage amp capable of 50-100W o/p at 500mV sens for full o/p = simplicity. However the noise performance isn't impressive. -70dB down/1w ( a good HiFi LS will pick this hiss up to the ear closeby)
I find all these phasesplitter types suffer from rising frequency misbalance with thd but for MI that is least important. Other users perhaps use a med u triode freestanding phasesplitter and then use a front end stage for gain and nfb applied here. Your choice.
The LTPconfig is usually DC coupled to the 1st stage. There are winnners and loosers in this, the winner, stability can be upped, more global nfb applied, 30dB typ; but I find the sound is more muddled.The looser is with such high feedback, bass transient overhang an issue as with so much global nfb the recovery is poorer. The Mullard 20W amp is a typ example.The noise floor is brilliant (aka -95dB/w); 30dB global nfb sounds clinical; But with such high global nfb I cannot "slap" guitar bass on this amp without severe intermodulation interaction if another signal is applied to the other input.
The Linear 50W amp of the 1960's had an identical design, >Ok for single instrumentation but not for mixed signals. My favourite for sonic accuracy for both MI and HiFi is the concertina or split load phasesplitter.
Critics Welcome!!
richy
I find all these phasesplitter types suffer from rising frequency misbalance with thd but for MI that is least important. Other users perhaps use a med u triode freestanding phasesplitter and then use a front end stage for gain and nfb applied here. Your choice.
The LTPconfig is usually DC coupled to the 1st stage. There are winnners and loosers in this, the winner, stability can be upped, more global nfb applied, 30dB typ; but I find the sound is more muddled.The looser is with such high feedback, bass transient overhang an issue as with so much global nfb the recovery is poorer. The Mullard 20W amp is a typ example.The noise floor is brilliant (aka -95dB/w); 30dB global nfb sounds clinical; But with such high global nfb I cannot "slap" guitar bass on this amp without severe intermodulation interaction if another signal is applied to the other input.
The Linear 50W amp of the 1960's had an identical design, >Ok for single instrumentation but not for mixed signals. My favourite for sonic accuracy for both MI and HiFi is the concertina or split load phasesplitter.
Critics Welcome!!
richy
It is great for a solo guitar!
But Fender went further modifying it to get more gradual coloration additions with loudness.
I will search for that concertina PS as I did for others. I think I will use a long tail pair and a floating paraphase in my future amp, with some relays for switching, probably, and a couple of 12AZ7. The fact is how will affect the sound, just listening to it. I don't care about measuring distortion with a scope or some similar device.
I'm really satisfied with all the replies!
BTW, some friend told me old 25W Twins sounded great! Must be a combination between floating paraphase and self biased push pull output.
I'm really satisfied with all the replies!
BTW, some friend told me old 25W Twins sounded great! Must be a combination between floating paraphase and self biased push pull output.
It is great for a solo guitar!
It is also great for Hi-Fi!
I spotted the floating or anode follower phase splitter in a modern Japanese amp, i breadboard it and found out it sounds great; transparent, no tubey at all, really hifi sound. I was so impressed i changed my pp amp splitter and no regrets.
Well one shortcoming is that i can't hear anymore crappy recordings, too painful.
And there are a lot!
It is also great for Hi-Fi!
I spotted the floating or anode follower phase splitter in a modern Japanese amp, i breadboard it and found out it sounds great; transparent, no tubey at all, really hifi sound. I was so impressed i changed my pp amp splitter and no regrets.
Well one shortcoming is that i can't hear anymore crappy recordings, too painful.
And there are a lot!
Last edited:
Thanks:....: I see it has the bit about sliderules in Leo's history....For those of us left who used'em is a great touch. Now wer'e in real thinking business.
www.aikenamps.com also has sim phasesplitter stuff.
richy
www.aikenamps.com also has sim phasesplitter stuff.
richy
Guys,
I have built quite a few guitar amps. Most recently I built a smallish low power, 2 channel amp for myself for home use and to cart around to friends houses for Sunday afternoon jam sessions etc.
I decided early to make it an all octal tube amp - just to set it aside visually from others and also decided to try out a few other things. I had always used the "stacked schmidt" splitter for guitar amps, as do 90+% of all amps I've ever seen. This time I decided to try a concertina splitter using a 6SL7 driving a pair of 6V6. The 6SL7 concertina is just lovely for guitar. I posted the schematic on PowerScaling.com. In a response to my post, Kevin O'Connor let slip that he always uses concertina splitters in all of his amps and that (for guitar use) the concertina sounds best when using a higher mu tube. So a 12AT7 or a 6SL7 is preferred. This is counter what is recommended when using a concertina splitter for HIFI, in that case use of low mu tube (12AU7, ECC88 etc) is recommended.
This amp has turned out be just stunning with a lovely overdrive. I will be using concertina splitters in all new guitar amp builds.
CAVEAT: The 2 channel preamp is all 6SL7 and the output tubes (6V6) are individually CCS biased with bypass caps - like my Baby Huey HiFi Amps - so it is NOT just the concertina splitter contributing to the tone and overdrive characteristic BUT it certainly helps. I also fitted an Ultralinear/Pentode Mode Switch. This switch is switched to the Pentode Mode position and stays there, so that switch was a waste of space.
Cheers,
Ian
I have built quite a few guitar amps. Most recently I built a smallish low power, 2 channel amp for myself for home use and to cart around to friends houses for Sunday afternoon jam sessions etc.
I decided early to make it an all octal tube amp - just to set it aside visually from others and also decided to try out a few other things. I had always used the "stacked schmidt" splitter for guitar amps, as do 90+% of all amps I've ever seen. This time I decided to try a concertina splitter using a 6SL7 driving a pair of 6V6. The 6SL7 concertina is just lovely for guitar. I posted the schematic on PowerScaling.com. In a response to my post, Kevin O'Connor let slip that he always uses concertina splitters in all of his amps and that (for guitar use) the concertina sounds best when using a higher mu tube. So a 12AT7 or a 6SL7 is preferred. This is counter what is recommended when using a concertina splitter for HIFI, in that case use of low mu tube (12AU7, ECC88 etc) is recommended.
This amp has turned out be just stunning with a lovely overdrive. I will be using concertina splitters in all new guitar amp builds.
CAVEAT: The 2 channel preamp is all 6SL7 and the output tubes (6V6) are individually CCS biased with bypass caps - like my Baby Huey HiFi Amps - so it is NOT just the concertina splitter contributing to the tone and overdrive characteristic BUT it certainly helps. I also fitted an Ultralinear/Pentode Mode Switch. This switch is switched to the Pentode Mode position and stays there, so that switch was a waste of space.
Cheers,
Ian
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.