‘I wanna to buy’ stuff is very good value
Hi
‘I wanna to buy’ stuff is very good value I have built 3 amps using the kit you have shown. As long as you use good Toroidal (I buy mine from RS
Transformers | Transformers | Toroid Transformers
And use quality alp pots
You can but these from Rapid electronics UK
Very cheaply.
http://www.rapidonline.com/sku/Elec...rofessional-audio-potentiometer/65228/66-0225
Build the chassis and screen everything careful and they sound amazing. I have had several friends with golden ears who are amazed at the quality for price. I have also built (rebuilt a valve preamp supplied by 'I wanna buy' for £18 which sounds really nice driving some Peter Daniels kits. There seem to be a bit of snobbery about Chinese products practically everything is made in china these days. I am currently looking at a remote control kit. If like me you’re strapped for cash this is a great way to experiment and put together a descent system.
Go for it and
Have Fun
Hi
‘I wanna to buy’ stuff is very good value I have built 3 amps using the kit you have shown. As long as you use good Toroidal (I buy mine from RS
Transformers | Transformers | Toroid Transformers
And use quality alp pots
You can but these from Rapid electronics UK
Very cheaply.
http://www.rapidonline.com/sku/Elec...rofessional-audio-potentiometer/65228/66-0225
Build the chassis and screen everything careful and they sound amazing. I have had several friends with golden ears who are amazed at the quality for price. I have also built (rebuilt a valve preamp supplied by 'I wanna buy' for £18 which sounds really nice driving some Peter Daniels kits. There seem to be a bit of snobbery about Chinese products practically everything is made in china these days. I am currently looking at a remote control kit. If like me you’re strapped for cash this is a great way to experiment and put together a descent system.
Go for it and
Have Fun
I also think the Chinese are getting a lot of stick for nothing. There are bound to be a few freudsters out there, but there are everywhere. I have just connected up my Goldmund clone without a case as yet, but I have music. I had a bit of smoke at first. I think it might have been because I had no input or speakers connected. I have to adjust the pots for the dc offset and bias and I assumed it was done without input or speakers. Seems I was wrong. I'm not very technical. I also used silver solder (4.7%) and I found it very difficult. You can't see easily if there is a dry joint. Seems I will need a buffer of some sort so I am looking at the Pass B1 clone on ebay. So far so good. Can't wait to compare it to the gainclone. I can only say it sounded good for the few minuets I tested it for although the heatsink got to hot to touch.
buy 63/37 solder.
This eutectic is very easy to use. Has near the minimum melting point of all the electrical solders.
With a good flux it wets very well.
It is cheap.
It is available.
Do not buy 60/40 nor 40/60. They are both unsuitable.
what size is your heatsink?
What is your ambient temperature.
I doubt the "too hot to touch" is right. I'll go further, it is wrong.
Hi Andrew, thanks for the help with the solder. I'm afraid I have no idea what the ambient temperature is on the Goldmund. I used a heat sink from ebay 200mm long with about 6 fins. I thought it might be too small, but I am going to use a low profile case is in the earlier pics of my Yuanjing amp a few posts back. Can someone tell me if I measure the DC offset with speakers connected and an input connected. PLease help. I think I must have offended Linuxguru as I have asked him too many questions. Although technically the Goldmund clone does not belong to this thread, there is a strong link as both Linuxguru and myself have made the Yuanjing amp and we are both constructing the Goldmunds' and it will be very interesting to compare results with each other and with the Yuanjing boards. I think the gainclones are hard to beat if put together well. I have sold off both my John Shearne Phase 2 and my Naim Nait 3 as they are no match. It will be interesting to see if this Goldmund containing 30 transistors and many other parts plus a pre-amp to really intensify the signal path can beat simplicity. I'm told the circuit is one of the best so I am very excited to make the comparison. Other members such as David Hugo, skodaboy, and borninthefifties have also made or bough assembled these Yuanjing boards which include a op-amp for starters plus power supply for £9. I don't know how they do it 'fake' or not.
I'm not offended - just preoccupied with other matters, while recovering from cold/cough for the last 10 days. My Goldmund kit has arrived, but I have not been able to look at it or check for missing/additional components, etc.
Regarding the outputs on the Goldmund running too hot, it's most probably an incorrect setting of the Class-AB output stage quiescent current. It should be set to between 50 - 100 mA by adjusting the potentiometer in the Vbe multiplier. You can measure the quiescent current by measuring the DC voltage across the emitter ballast resistor of each O/P transistor and dividing it by the value of the resistor. For example, 0.2 V / 0.22 ohms ~= 90 mA Iq.
You need to set the DC offset initially without the speakers connected. You can perform a fine re-adjustment after 1 hour of warm-up with the speakers connected.
Regarding the outputs on the Goldmund running too hot, it's most probably an incorrect setting of the Class-AB output stage quiescent current. It should be set to between 50 - 100 mA by adjusting the potentiometer in the Vbe multiplier. You can measure the quiescent current by measuring the DC voltage across the emitter ballast resistor of each O/P transistor and dividing it by the value of the resistor. For example, 0.2 V / 0.22 ohms ~= 90 mA Iq.
You need to set the DC offset initially without the speakers connected. You can perform a fine re-adjustment after 1 hour of warm-up with the speakers connected.
Thank you Linuxguru, I will make the adjustments once I have boxed it. Hope your cold is better. When you look at the contents I would be grateful if you could tell me if you have 2 brown saha capacitors. I don't seem to have anywhere to mount them. They could just be in the package by mistake off course. I emailed the supplier but he never answered. Good luck. Regards Barry
Another set of mods to the Jim's Audio (nearly the same as the Yuanjing) buffered gainclone, which gives subtle but audible improvements to the sonics:
1) Change the gain of the LM3886 from 21/23 (feedback resistor of 20k/22k) to a higher value - say 28 or 34 (feedback resistor of 27k or 33k). I went with 27k on my second build, and the sonics are subtly better than the first build (with 22k). Also change the input resistor to match the feedback resistor - I'm getting output offset voltages of 0.5 and 0.3 mV in the two channels (unbelievably low, but obtainable if you match the input resistor and the feedback resistor to within 1%).
2) Use Allen-Bradley or similar carbon-composition resistors for the feedback resistor and input resistor of the LM3886. With metal-film resistors at these locations, there was a tinny/ringing artifact in the upper-mids which has completely vanished with the Allen-Bradleys. The tradeoff is that there is now a slight audible hiss at quiescence, because of the higher noise figure of the carbon-composition resistors. YMMV, but that's a tradeoff that I'm comfortable with. "Light My Fire" by The Doors was completely unlistenable earlier (excessively tinny), but is acceptable now. Likewise, many ABBA numbers which are shrill at the source.
3) I changed the Boucherot to 10 ohms, 1W metal-oxide in series with a 100nF polyester. Unsure of the impact on stability, but there's no negative impact on sonics.
4) Small electrolytics on +Vs and -Vs changed from 680 uF Rubycon JXA to 150 uF Panasonic FJ (as suggested in other threads) - detail seems to have improved. It's probably worth experimenting with 220 or 330 uF also.
5) Main PSU filter cap reduced to 4700uF, 50V Elna CEW from 6800uF, 50V Panasonic TSU. No appreciable impact on sonics.
1) Change the gain of the LM3886 from 21/23 (feedback resistor of 20k/22k) to a higher value - say 28 or 34 (feedback resistor of 27k or 33k). I went with 27k on my second build, and the sonics are subtly better than the first build (with 22k). Also change the input resistor to match the feedback resistor - I'm getting output offset voltages of 0.5 and 0.3 mV in the two channels (unbelievably low, but obtainable if you match the input resistor and the feedback resistor to within 1%).
2) Use Allen-Bradley or similar carbon-composition resistors for the feedback resistor and input resistor of the LM3886. With metal-film resistors at these locations, there was a tinny/ringing artifact in the upper-mids which has completely vanished with the Allen-Bradleys. The tradeoff is that there is now a slight audible hiss at quiescence, because of the higher noise figure of the carbon-composition resistors. YMMV, but that's a tradeoff that I'm comfortable with. "Light My Fire" by The Doors was completely unlistenable earlier (excessively tinny), but is acceptable now. Likewise, many ABBA numbers which are shrill at the source.
3) I changed the Boucherot to 10 ohms, 1W metal-oxide in series with a 100nF polyester. Unsure of the impact on stability, but there's no negative impact on sonics.
4) Small electrolytics on +Vs and -Vs changed from 680 uF Rubycon JXA to 150 uF Panasonic FJ (as suggested in other threads) - detail seems to have improved. It's probably worth experimenting with 220 or 330 uF also.
5) Main PSU filter cap reduced to 4700uF, 50V Elna CEW from 6800uF, 50V Panasonic TSU. No appreciable impact on sonics.
I have just performed some of Linuxguru's mods. Changed the gain and feedback and input resistors. Changed small electrolytic. I had already added 2 x Elna 4700uf filter caps because I used a seperate power board to keep the ac away. I tried removing the original caps on the Yuanjing but it made no audible difference that I could hear so put them back just to make the board look better. The other suggested mods have certainly improved things. Like Linuxguru, I played my ABBA cd's and two women screatching at the top of their ranges is a good test for shrill and the improvement is quite noticable. I haven't tried lifting the ground again. I'm afraid of what might happen. Just connected the Goldmund again for a few minuits and it is a very different sound, more subtle, like my old quad valve amp. Oh yes, I think i'm gonna like this. Played a bit of Beethoven's Egmont and a bit of Floyd. Can't wait to get it all fixed up in a case. God, I wish I was a millionaire!
and if you were, you would pay someone else to prebuild and install all your gear.
Where would be the enjoyment of creating your own DIYaudio?
I'm awaiting the arrival of LM4562 opamps for the Jim's Audio (similar to Yuanjing) 2 x LM3886 boards. Meanwhile, I rolled and auditioned NE5532, TL072, JRC4560, JRC4580, LT1208 and LT1213.
Results: The best overall value is still the NE5532 - a great all-round performer across the band, and costs all of $0.20. However, for detail, the LT1213 seems to be better, although a trifle bright in the upper-mids/treble.
The earlier change of the feedback resistor from 22k to 27k (carbon-composition) was a more favourable change in sonics than all of the op-amp trials. The LM833 has also been suggested as a replacement for the NE5532; but for now, I'm keeping the LT1213 until the LM4562 shows up and proves itself.
Edit: I'm using an ST D25SBA80 (25A, 800V) bridge-rectifier in board #2 in place of a HY GBJ810 8A,1000V unit in board #1. It has not hurt the sonics in any way, while being far more durable. I recommended the 25A device if you can get it - the price premium over the 8A device was only about 20%. Are there any fast, soft-recovery devices that are recommended here?
Results: The best overall value is still the NE5532 - a great all-round performer across the band, and costs all of $0.20. However, for detail, the LT1213 seems to be better, although a trifle bright in the upper-mids/treble.
The earlier change of the feedback resistor from 22k to 27k (carbon-composition) was a more favourable change in sonics than all of the op-amp trials. The LM833 has also been suggested as a replacement for the NE5532; but for now, I'm keeping the LT1213 until the LM4562 shows up and proves itself.
Edit: I'm using an ST D25SBA80 (25A, 800V) bridge-rectifier in board #2 in place of a HY GBJ810 8A,1000V unit in board #1. It has not hurt the sonics in any way, while being far more durable. I recommended the 25A device if you can get it - the price premium over the 8A device was only about 20%. Are there any fast, soft-recovery devices that are recommended here?
I tried various op-amps but I really could't notice much improvement but left it with the LM4562. I think it was a little bright but your resistor mod seems to have rectified that and there is now good detail and soundstage. I have ordered a nice cheap 19" rack case from Jaycar Electonics for about £35 for the Goldmund. I am going to Scotland on Tuesday to bag a few Munro's. I hope the case will be here when I get back (if I don't fall off a mountain!).
4x100watt for $50 !
This is nuts !
4*100 watt @ 4ohm, TK2050 D-class Audio Amplifier Board - eBay Power Amplifiers, Amplifiers, Home Audio, Electronics. (end time 12-Mar-10 22:53:56 AEDST)
Very tempted...
This is nuts !
An externally hosted image should be here but it was not working when we last tested it.
4*100 watt @ 4ohm, TK2050 D-class Audio Amplifier Board - eBay Power Amplifiers, Amplifiers, Home Audio, Electronics. (end time 12-Mar-10 22:53:56 AEDST)
Very tempted...
Linux guru thanks for the opamp review.Its very useful for those starting out who dont have a stock parts to try.I have ordered opa264 to try but if theres not much difference ill save them (electronic crossover?).Why is there such a price jump from ne5532 to other audio opamps?!
I have the input resistors, input cap and lifted ground mods done ,fired it up (on old speakers) no probs .no hum only some noise at full volume with ear next to tweeter. Haning final dreesing taken out of my ear on onday so hope to give a good listening test with restored hearing!even downloaded some ABBA to test the ringing effect.
Finally whya cases so expensvie compared to the electronics bits. Idont want to spend $50 on postage for a tin box!
I have the input resistors, input cap and lifted ground mods done ,fired it up (on old speakers) no probs .no hum only some noise at full volume with ear next to tweeter. Haning final dreesing taken out of my ear on onday so hope to give a good listening test with restored hearing!even downloaded some ABBA to test the ringing effect.
Finally whya cases so expensvie compared to the electronics bits. Idont want to spend $50 on postage for a tin box!
There's not much difference - in fact, faster op-amps often sound worse (more tinny) in this board, which I suspect is because of poor ground layout and EMI/RFI issues, which faster opamps are more susceptible to. NE5532 is probably cheap because it's been around for a long while and all development expenses have been recovered long ago. Besides, it's a very high-volume part.
It looks like your lifted ground mods are done correctly. It should cure most (say 95%), but not all of the hum/hoise - some residual hum can be heard by keeping your ear next to the drivers, as you did. It's not annoying or particularly noticeable. I conjecture that some of this hum can also be controlled by using a fast, soft-recovery rectifier for the main bridge rectifier. I get excellent BYV-32E TO-220 rectifiers here, but they're not bridge rectifiers. So I ended up using a regular (slow-recovery) ST D25SBA80 25A, 800V rectifier instead.
Alternatively, use a separate snubberized power-supply board with the large filter capacitors on it, and feed only low-pass filtered DC to the gainclone board. YMMV, but this increases the cost and complexity of the setup a bit.
Yup, I noticed that EBay prices for quality amplifier cabinets are exorbitant. To be fair, they are well-made, and intended for high-end/boutique equipment, where the price of the connectors often exceeds the price of the box. Moreover, the cost of shipping the cabinet usually kills the deal.
Hi Skodaboy, The Chinese case I bought for the Yuanjing board was expensive. The postage was more than the case, but to be fair it is excellent and comes pre-drilled and with feet, aluminium knobs, on-off switch, mains socket, fuse holder, and very good quality phono sockets. The finish is great and makes the amp look like something commercial. I looked everywhere for something cheaper for the Goldmund and was advised to look at Jaycar Electronics who are not just in the UK but also in Australia and New Zealand so could be useful. This is a 19" 2U rack mount kit that comes complete with all panels (unlike the RS versions) and even handles. It is priced at about £35 including postage in the UK so should be about the equal in New Zealand. I hope this helps.
Last edited:
Another set of mods (yeah, I'm getting obsessed with improving this cheap amp), extending the lifted-ground concept to the LM3886 section. The aim of this mod is to greatly reduce the effect of ground bounce (in particular, the ground return current from the speakers) on the small-signal resistor-networks near the LM3886, which are grounded at two locations in the middle of the return plane. This mod lifts those two grounds and jumpers it to the previously-created lifted ground of the NE5532 preamp/buffer.
Caution: this mod will increase the output offset at the LM3886 output, which I measured to be between 100 mV - 200 mV. If this bothers you, *stop right here* and don't attempt this mod!
Pre-requisite: The previous lifted-ground mod for the NE5532 preamp/buffer.
1) Remove the DC-blocking capacitor in the voltage-series feedback network of the LM3886. This is shown to be 22 uF, 50V on the silkscreen.
2) Lift the grounded end of the input shunt resistor (22k or 27k as the case may be) and bridge it to the adjacent 1k resistor of the voltage-series feedback network. This node is the newly-created lifted-ground for the LM3886.
3) Jumper the lifted-ground of the LM3886 to the lifted ground of the NE5532 with an insulated piece of hook-up wire on the solder side of the board.
The pictures attached below show the mods for one channel:
Caution: this mod will increase the output offset at the LM3886 output, which I measured to be between 100 mV - 200 mV. If this bothers you, *stop right here* and don't attempt this mod!
Pre-requisite: The previous lifted-ground mod for the NE5532 preamp/buffer.
1) Remove the DC-blocking capacitor in the voltage-series feedback network of the LM3886. This is shown to be 22 uF, 50V on the silkscreen.
2) Lift the grounded end of the input shunt resistor (22k or 27k as the case may be) and bridge it to the adjacent 1k resistor of the voltage-series feedback network. This node is the newly-created lifted-ground for the LM3886.
3) Jumper the lifted-ground of the LM3886 to the lifted ground of the NE5532 with an insulated piece of hook-up wire on the solder side of the board.
The pictures attached below show the mods for one channel:
Attachments
Last edited:
After this mod (lifted ground for the LM3886), the LT1208 as the pre/buffer easily trumps the LT1213 (and most probably all the other op-amps previously tested). There's a certain crystalline transparency and fluidity which wasn't there earlier. Without the mod, it's too bright and screechy.
I'll await the LM4562 - it probably won't disappoint. A more expensive (and highly touted) alternative is the THS4032 - probably overkill to use $20 worth in opamps/adapters in a $15 board...
I'll await the LM4562 - it probably won't disappoint. A more expensive (and highly touted) alternative is the THS4032 - probably overkill to use $20 worth in opamps/adapters in a $15 board...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.