Trying to make sense of the Sweet Peach - Page 22 - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 18th March 2012, 11:13 PM   #211
diyAudio Member
 
Globulator's Avatar
 
Join Date: Aug 2004
Location: Cambridge, England.
The different cathode resistors were just experiments in the bias current of the tube, which affects which part of the tube curves the audio uses.

Read up about tube load lines and look at the tube data sheets.
The 12AU7 needs heater re-wiring for it, so is not a direct swap.
People have found the 6n6p works well in there (just plug it in) so I'd try that.

Personally I did not find the curves looked that good though, but I didn't ever try it so have a go - it's much cheaper and easier than a AU7.

If you have already for the AU7 in (Tung-Sol are good!) then I'll have to have another look at the diagram - but 1.5k sounds a safe bet.

There are really 2 stages to the mods:

1) Change the caps etc and try a 6N6p or rewire heaters for an AU7.
2) Rebuild it completely.

Eventually I ended up at 2) - documented in the last few pages, and now I may even tweak that a little with some later ideas.
  Reply With Quote
Old 18th March 2012, 11:26 PM   #212
morme is offline morme  United States
diyAudio Member
 
Join Date: Nov 2004
Location: ojai
Default Cathode resistor changes

I made the 12AU7 swap (rewired for 6V w/ cut trace which was no big deal with a dremel.)

I chanced the 12AU7 cathode resistor to 2.2k as recommended in your post 58

I was just wondering what differences you heard changing the 6N3 cathode resistors from 1k (stock) to 2.2k. Changing to 1.5k can be easily done if the change to 2.2k was a mistake.

I am not really up for rebuilding the entire amp unless I rewired it into pentode mode but I was told that the transformers are not up to 18 watt output.

I could just make the resistor swap and listen, but sometimes these things are so subtle that I am not sure what I am hearing.

thanks again


Quote:
Originally Posted by Globulator View Post
The different cathode resistors were just experiments in the bias current of the tube, which affects which part of the tube curves the audio uses.

Read up about tube load lines and look at the tube data sheets.
The 12AU7 needs heater re-wiring for it, so is not a direct swap.
People have found the 6n6p works well in there (just plug it in) so I'd try that.

Personally I did not find the curves looked that good though, but I didn't ever try it so have a go - it's much cheaper and easier than a AU7.

If you have already for the AU7 in (Tung-Sol are good!) then I'll have to have another look at the diagram - but 1.5k sounds a safe bet.

There are really 2 stages to the mods:

1) Change the caps etc and try a 6N6p or rewire heaters for an AU7.
2) Rebuild it completely.

Eventually I ended up at 2) - documented in the last few pages, and now I may even tweak that a little with some later ideas.
  Reply With Quote
Old 18th March 2012, 11:47 PM   #213
diyAudio Member
 
Globulator's Avatar
 
Join Date: Aug 2004
Location: Cambridge, England.
Quote:
Originally Posted by morme View Post
I made the 12AU7 swap (rewired for 6V w/ cut trace which was no big deal with a dremel.)

I chanced the 12AU7 cathode resistor to 2.2k as recommended in your post 58

I was just wondering what differences you heard changing the 6N3 cathode resistors from 1k (stock) to 2.2k. Changing to 1.5k can be easily done if the change to 2.2k was a mistake.

I am not really up for rebuilding the entire amp unless I rewired it into pentode mode but I was told that the transformers are not up to 18 watt output.

I could just make the resistor swap and listen, but sometimes these things are so subtle that I am not sure what I am hearing.

thanks again
You need to find out the voltage supply to the tube and the anode resistor, as I can't recall. If it's running at 380V with a 47K resistor then to balance the signal (i.e. midpoint) you need (380/2) / 47000 = 4mA, the ECC82 datasheet shows you need a -7.5V grid voltage.

So at 4mA that's 7.5 / 0.004 = 1875 ohms needed for the cathode resistors (2k). Can you measure the voltage from the anode resistor to ground and see what it is? I'm guessing 380V. Also measure the anode resistor as I am going from memory here.

Then use the calculations above and the ECC82 datasheet to see what you get. For 380V and 47k the cathode resistor is ideally 2k though.

The pentode mod requires metalwork etc - it's a very big rebuild, but the transformers handle the power well. It's less than 18W though
  Reply With Quote
Old 19th March 2012, 06:42 PM   #214
morme is offline morme  United States
diyAudio Member
 
Join Date: Nov 2004
Location: ojai
Default 6N3 Cathode resistor change, cont.

These recommendations are for the 12AU7 valve I assume.

I was wondering what differences you heard when you changed the cathode resistor to 2.2k on the 6N3 valve? Is there a good replacement for that tube?

What do you call the mod described in post 58? I have made all the changes there except replacing the WIMA box caps. Replacements are still "in the mail"

Any other recommendations (leaving the feedback circuit basically intact)?

I havn't had a chance to make any voltage measurements. I'll do that soon for all 5 tubes.

Quote:
Originally Posted by Globulator View Post
You need to find out the voltage supply to the tube and the anode resistor, as I can't recall. If it's running at 380V with a 47K resistor then to balance the signal (i.e. midpoint) you need (380/2) / 47000 = 4mA, the ECC82 datasheet shows you need a -7.5V grid voltage.

So at 4mA that's 7.5 / 0.004 = 1875 ohms needed for the cathode resistors (2k). Can you measure the voltage from the anode resistor to ground and see what it is? I'm guessing 380V. Also measure the anode resistor as I am going from memory here.

Then use the calculations above and the ECC82 datasheet to see what you get. For 380V and 47k the cathode resistor is ideally 2k though.

The pentode mod requires metalwork etc - it's a very big rebuild, but the transformers handle the power well. It's less than 18W though
  Reply With Quote
Old 19th March 2012, 08:52 PM   #215
diyAudio Member
 
Globulator's Avatar
 
Join Date: Aug 2004
Location: Cambridge, England.
Hi Morme,

The mod doesn't have a name so we'll have to call it Mod58 !
Yes my comments are for the ECC82, datasheet attached.

The biggest change (after the caps) was the ECC82, much better than the old mis-biased 6N1p.

I tried LED biasing but didn't like the sound, I was trying quite a few things. After a while the 6N3p tubes only being half used got to me, and Wavebourn told me they were poor for audio, and they are wired up in a weird way making substitution difficult so I redesigned the whole thing with new tubes and a much (much) shorter feedback path.

The main issue with the Mod58 is that it still uses the global feedback loop and the GU50 in triode mode, switching to pentode with proper feedback was a big jump.
Attached Files
File Type: pdf ecc82-philips1954.pdf (185.6 KB, 33 views)
  Reply With Quote
Old 19th March 2012, 10:56 PM   #216
morme is offline morme  United States
diyAudio Member
 
Join Date: Nov 2004
Location: ojai
Señor Globulator, Do you by chance have a schematic of your final pentode version?

thanks for all your help

btw, I just ordered a new 12AU7A Amperex Globe Logo NOS made in JAPAN on Mullard tooling ribbed plates in original boxes 1970's. It's supposed to be a good sounding tube.

m

Quote:
Originally Posted by Globulator View Post
Hi Morme,

The mod doesn't have a name so we'll have to call it Mod58 !
Yes my comments are for the ECC82, datasheet attached.

The biggest change (after the caps) was the ECC82, much better than the old mis-biased 6N1p.

I tried LED biasing but didn't like the sound, I was trying quite a few things. After a while the 6N3p tubes only being half used got to me, and Wavebourn told me they were poor for audio, and they are wired up in a weird way making substitution difficult so I redesigned the whole thing with new tubes and a much (much) shorter feedback path.

The main issue with the Mod58 is that it still uses the global feedback loop and the GU50 in triode mode, switching to pentode with proper feedback was a big jump.
  Reply With Quote
Old 31st March 2012, 09:24 PM   #217
XavLNK is offline XavLNK  Belgium
diyAudio Member
 
Join Date: Dec 2010
Location: Romsée
Default Some audio measurements

Hello all,

I recently got a defective Sweet Peach from a friend. It soon appeared that the sockets of the 6N1 and 6N3 are not of highest quality. Moving a bit the 6N3 in its socket solved the problem... But the quality of the whole stuff is not so fantastic, to say the least. It is a cheap product so let's not expect miracles.
By the way, no offense for the Chinese that are capable of building real jewels.

Still... As that the amplifier was fixed, I listened a bit to it. The first impression was that the sound was comparable to the inside of the amplifier. Really not impressive. Even disappointing.

Then, I made some reverse engineering (on a chinese design ) to figure out what could be improved. So I did the following:
- Replace the 0.1uF input cap (supposed to be Wima) by high quality Panasonic 1uF.
- Replace the 0.1uF/250V (also Wima) by high Quality Audyn Plus caps.
- Replace the 4 cathode resistor decoupling electrolytic capacitors (220uF/35V) by Elna SILMIC II models with same ratings (that's all I had in stock), with 10nF MKP in parallel.
- Bypass the 220uF/35V electrolytic capacitor in series with the feedback network.
- Rebias the 2 FU-50 to 85 mA each (they were completely off: one at 90 mA and the other one at 55mA).

Click the image to open in full size.
Click the image to open in full size.

Listen again... At least I started to get a sense of a realistic sound stage and decent sound balance. But not yet to where I wanted to go. I was expecting these changes to have a lot more effects.

Then I fortunately found that thread and the post about the hidden 0.47uF capacitor on the POT / Selector board. I immediately bypassed these 2 caps and listened again.
Wow, what a difference now . It is becoming pretty interesting.

Note that I compare this little amp to a Cayin 500 with NOS tubes and to a heavily mod'ed 300B SE amp with modern Shuguang tubes (I found it on a second hand street market in Hong Kong...).

I also tried the 6N6 instead of the 6N1. The difference is not so obvious. I might have a slight preference for the 6N6 but not on all tracks. Female voice are a bit more pleasant with the 6N6. But the overall feeling of the amplifier of being in control of the speakers is not so great.

The gain of the 6N6 is lower than the one of the 6N1. The feed back loop does not play a significant role anymore and the output impedance of the amplifier is increasing significantly (though I did not measure it). That might explain what I heard.

I made some measurements of the amplifier with an Audio Precision analyzer, to see what I had in my hands.

First, I measured the effective power that could be obtained before the distortion shape changes dramatically. I obtained 6 Watts per channel.
If you accept to go up to 2% THD, the power is around 8 Watts.
With a speaker efficiency >90 dBSPL (mine are 91dBSPL), it is not an issue at all.
Preferably use a speaker impedance of 8 Ohms.

I made some THD measurements (I know that I should not talk about that when dealing with tubes, but I did...). I am giving all the plots for reference.

The frequency response
With the 6N1, it extends well over 20 kHz. The LF -3dB is around 5 Hz.
The 6N6 shows a loss of gain of about 2 dB (I tried 2 models, a chinese 6N6 from the same manufacturer than the 6N1 and NOS russian 6N6P). The treble extension is also slightly reduced (less feed back...). The bass goes 5 Hz lower.

Click the image to open in full size.

The THD+N versus Frequency
I made the measurement for various FU50 bias level, from 50 mA to 100 mA.
I did not measure above that because the plate dissipation would be exceeded (40W) so nobody would want to use the amplifier like that.
I took the measurements at a power of 3 Watts, in a load of 4R7 with the impedance switch set to 4R.
The bias at 50 mA is really not good in terms of THD (2%).
The bias at 100 mA is decent at 0.6% but not much better that the bias at 80 mA at 0.7%. The H3 harmonics is getting more present with 100 mA.
I found 85 mA to be a very acceptable trade-off in terms THD vs power consumption.
With this bias, the sound stage is relatively accurate and stable.
With 85 mA, H3 is 20 dB lower than H2.
The big disappointment of that amplifier is the THD below 200 Hz.
It is increasing very fast when the frequency goes low. I suspect the output transformer to be responsible for that. That probably explains why I do not find this amplifier as good sounding as my 300B SE that has about the same output power (but OPT 4 times as big).

Click the image to open in full size.

THD versus power
Not much to say, here.
THD increases regularly with power, which is normal. Still a bit high for my likings. You can see the 6W limits. I did the measurement with the 6N1 and the 6N6. The 6N1 has better THD figures because of the NFB playing a bigger role.

Click the image to open in full size.

I did not change anything in the topology of the amplifier (it is not mine, after all). With better 6N3 and 6N1 (NOS russians?), the overall sound quality may improve. As I do not have these tubes available, I can't tell.

Bus, as such, with the 6N1 or the 6N6, the slightly modified Sweet Peach outperform most average transistor amplifiers I have heard.
For $300, we definitively can't complain.

Have fun with it.
  Reply With Quote
Old 6th April 2012, 10:39 AM   #218
diyAudio Member
 
Join Date: Mar 2012
I just replaced the Chinese tubes with (matched) Russian 6N3P. They've not been in there for half an hour yet and even now it seems mid and high opened up. Is this possible or is this just a psychological trick my brain plays on me?
  Reply With Quote
Old 9th April 2012, 01:24 AM   #219
861147 is offline 861147  United States
diyAudio Member
 
Join Date: Apr 2012
Default Phono stage?

Hi all!

I think it is great that we have discovered this cheap Chinese tube amp that with minor mods, we can make it sing as well as $2000+ tube amps. The reason I'm considering this particular amp is that it has a phono stage built into it and I've been looking for an integrated tube amp with phono to save a bit of money and space! So, my question is, has anyone here tested the phono stage and does it sound at all decent?

Like a lot of Chinese amps, this one has major potential for around $300. I'm only a little worried about attempting some of the mods mentioned here, I have little hands-on electronic experience though I have read about the science behind these amps constantly, since I picked up my first tube amp several months ago. I now have one vintage Knight int. tube amp employing the 6GW8 power tubes (HARD to find), a small vintage Japanese pre-amp/phono stage with two 12AX7s (works although with obvious hum), and most recently, a new Chinese 6P1/6N2 amp ($250 total) that I thought would work with a turntable because of the high gain of the 6N2s (similar to 12AX7) and an input labeled "LP". I was wrong. I should've realized that it wouldn't work when there was no TT ground connection on the back. Anyway, I have my eye on this amp and this thread, now that I've found a tube amp that might actually do it all (TT, mp3, DAC, line-level amplification, etc.)

Thanks.
  Reply With Quote
Old 16th April 2012, 08:11 PM   #220
frapa02 is offline frapa02  United Kingdom
diyAudio Member
 
Join Date: May 2009
Location: Berkshire, UK
Following up from the comments by XavLNK on the performance, I concur with your thoughts about the limitations of the output transformer. Thanks for the extensive analysis, this is very interesting.

As per my comments in the other thread re distortion, if I just try to put more than a watt or two of sine wave through the EL34B peach below 100Hz you get serious transformer saturation. It is really very bad.

Also, I tried testing the amp using just the driver and output stage and it sounded like a guitar amp and looking at a 1kHz sine wave on the scope I could see a distorted wave shape at even moderate power levels. This was only occurring in the output stage and looked like 2nd harmonics. restoring the feedback and pre-amp tends to overcome most of this problem. I even tried swapping the EL34Bs for KT88s with no perceivable change, re-biased of course, keeping it around 55mA.

I am going to try it with some high quality transformers - I am waiting for a mate to lend me some. The problem is that the wiring is very hard to get to.

I can confirm that the primary of the transformers measures 5H inductance.

We can't expect more than this for the price and I am finding exactly what I would expect (and I wonder what is really hidden in the potting compound!).
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Linear voltage regulator: how to make good use of "sense" and "ground sense"? NeoY2k Analog Line Level 7 6th September 2008 11:35 PM
Does this make sense ? 00940 Solid State 7 27th March 2004 09:29 PM
does this make sense? nosegoblin Solid State 9 14th February 2004 02:19 PM
Does this make sense? Illusus Chip Amps 11 25th September 2003 01:59 PM


New To Site? Need Help?

All times are GMT. The time now is 05:54 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2