Interesting Tibi and Zsolt!
While you are optimising the Quad 405, I have revised my layout. Just a layman's work... I have dropped the DC-clamp in this one but is still without Tr3/R18.
http://www.twin-x.com/groupdiy/albums/userpics/Quad405_20080622_1840.GIF[{img]
Roger
While you are optimising the Quad 405, I have revised my layout. Just a layman's work... I have dropped the DC-clamp in this one but is still without Tr3/R18.
http://www.twin-x.com/groupdiy/albums/userpics/Quad405_20080622_1840.GIF[{img]
Roger
I cannot edit my post above. The link should read:
I have no idea if the layout is correct. I have followed Keith Snook's schematic "QUAD 405 - DCD Mod-3" and made a few changes. C15/C16 complemented with large value electrolytics, a couple of transistors à la Bernd Ludwig in the power supply lines to the IC. The DC-clamp is removed as many will use the Velleman protection circuitry. Fuses are Picufuse, directly soldered to the PCB. Multiple PCM for C1/C4/C2b/C6/C8/CD1/CD2 (5-10-15 mm etc.). CD1/CD2/C15/C16 could have the same PCM, 5 mm. The power resistors should be on the back of the PCB, no heating of C10 that way. D10/D11 could be soldered direcly on the Tr9/Tr10. Tr3 and R18 have been removed.
Roger
An externally hosted image should be here but it was not working when we last tested it.
I have no idea if the layout is correct. I have followed Keith Snook's schematic "QUAD 405 - DCD Mod-3" and made a few changes. C15/C16 complemented with large value electrolytics, a couple of transistors à la Bernd Ludwig in the power supply lines to the IC. The DC-clamp is removed as many will use the Velleman protection circuitry. Fuses are Picufuse, directly soldered to the PCB. Multiple PCM for C1/C4/C2b/C6/C8/CD1/CD2 (5-10-15 mm etc.). CD1/CD2/C15/C16 could have the same PCM, 5 mm. The power resistors should be on the back of the PCB, no heating of C10 that way. D10/D11 could be soldered direcly on the Tr9/Tr10. Tr3 and R18 have been removed.
Roger
tvicol said:
My advice is to use a dublet and to increase the gain to 5.
Tibi,
By that you mean Tr1b (2SA1085E) and Tr4 (ZTX753) without Tr3 (2SA1085E)?
I see that you have "opened the window" a bit more in your latest 405 MOSFET by increasing C6. Will it still be stable like that?
Roger
@RogerGustavsson
Nice PCB! Do you want me to make this pcb in Altium? I think if you increase the size of the pads for all components and it will be OK. Probably ExpressPCB has options to export the files in the format required by a factory.
Nice PCB! Do you want me to make this pcb in Altium? I think if you increase the size of the pads for all components and it will be OK. Probably ExpressPCB has options to export the files in the format required by a factory.
Zsolt,
That would be excellent! I will just try to make some adjustments the next few days. I will also tell the size of the holes where it is not clear. I do not like pins for connections. I have also started to look for some of the not so common components. They cannot be found at a single supplier.
I have been adjusting your layout made in Altium. Need a lot more traing.... ExpressPCB is a bit limited.
Roger
That would be excellent! I will just try to make some adjustments the next few days. I will also tell the size of the holes where it is not clear. I do not like pins for connections. I have also started to look for some of the not so common components. They cannot be found at a single supplier.
I have been adjusting your layout made in Altium. Need a lot more traing.... ExpressPCB is a bit limited.
Roger
@tvicol
@RogerGustavsson
For power and output I will use Faston FS1536 6.3mm solder terminals so one can solder to it or use a female push-on terminal.
Yes I know, by using Keith's version ~1.3V is needed for full output, but increasing the gain from 3.77 to 5 will increase the overall gain, and I would like to keep this ~1.3V - 1.5V for full output. Thats why I thought to further reduce the opamp gain.You don't need to reduce the gain of the opamp. This was already done by Keith.
@RogerGustavsson
For power and output I will use Faston FS1536 6.3mm solder terminals so one can solder to it or use a female push-on terminal.
RogerGustavsson said:
Tibi,
By that you mean Tr1b (2SA1085E) and Tr4 (ZTX753) without Tr3 (2SA1085E)?
I see that you have "opened the window" a bit more in your latest 405 MOSFET by increasing C6. Will it still be stable like that?
Roger
Roger,
I mean exactly like Keith did in his DCD-MOD4, only Tr4 and Tr7.
Tr1b is part of current source for Tr2.
In my MOSFET version C6 was reduced and bandwith increased over 100KHz and amplifier is rock stable. 🙂
Regards,
Tibi
vzs said:@tvicol
Yes I know, by using Keith's version ~1.3V is needed for full output, but increasing the gain from 3.77 to 5 will increase the overall gain, and I would like to keep this ~1.3V - 1.5V for full output. Thats why I thought to further reduce the opamp gain.
...
Zsolt,
In fact the class A gain will be increased from 3.77 to 6.
Overall amplifier gain will be:
[(22+3.3)/3.3]x[(500+100)/100]=46
Aprox 33.2dB
If you want to keep full output (100W/8ohm) for 1.5V you will need to decrease operational gain.
TIbi
I don't know if you guys will find this interesting as you are all far more knowledgeable than me but:
I added -11db attenuation to the input of my 405. Similar to some attenuating interconnects but done inside the amp. I can measure the resistor values if anyone would like to know them.
This really improved the sound, much clearer and far better treble performance. Of course combined with the fact that I have also reduced the gain to ~1.3v for full output as per DC Daylight mods, it means that the amp doesn't reach full volume, but it is still more than loud enough for me with the volume pot right at maximum.
Lee.
I added -11db attenuation to the input of my 405. Similar to some attenuating interconnects but done inside the amp. I can measure the resistor values if anyone would like to know them.
This really improved the sound, much clearer and far better treble performance. Of course combined with the fact that I have also reduced the gain to ~1.3v for full output as per DC Daylight mods, it means that the amp doesn't reach full volume, but it is still more than loud enough for me with the volume pot right at maximum.
Lee.
tvicol said:
Roger,
I mean exactly like Keith did in his DCD-MOD4, only Tr4 and Tr7.
Tr1b is part of current source for Tr2.
In my MOSFET version C6 was reduced and bandwith increased over 100KHz and amplifier is rock stable. 🙂
Regards,
Tibi
It seems that DCD-Mod-3 and DCD-Mod-4 have the same arrangement of the transistors (2SC2547E, two 2SA1085E and ZTX753) as well as most of the surrounding components, just that Keith have changed the name of Tr3/Tr4.
So raising the gain is the cure for oscillations in this case?
Roger
RogerGustavsson said:
...
So raising the gain is the cure for oscillations in this case?
Roger
Raising the gain will not cure oscillations, but will make the amplifier more stable.
To cure oscillations in this amplifier is a real challenge.
As I said here I didn't managed to get this amplifier stable with Darlington's. Mine oscillated at very high frequency.

Regards,
Tibi
I've promised a pcb for today, but I couldn't manage to finish it. I've started working on it but after 6months of not using Altium it takes some time to get accustomed to it again. So I need some more days... 😉
I have had several people tell me that links from DIYAudio to my site do not work - Initially this was due to the ISP changing to a case sensitive sever but as with the link posted at the beginning of this thread it is because the URL was ended with a "." period
I am concentrating more on the perfect RIAA amplifier and vinyl reproduction than power amps at present but as and when I get the time I'll update some more of the site
regards Keith
I am concentrating more on the perfect RIAA amplifier and vinyl reproduction than power amps at present but as and when I get the time I'll update some more of the site
regards Keith
I'm interested in making this modifications of Quad405.
Few days ago i opened my quad405 which i have made some 10 or 11 years ago. I saw that one out of two 560ohm resistors on both amps are almost turned to coil! So i bought some 5w to replace those old 2w. Flipping the board to replace resistors i saw the horror! The horror was my work before 10 years 🙂 The board was some cheap thing, has really thin connections which are pealing very easy! I saw traces of my previous modifications. I think i was replacing output transistors.
So, to get to the point: I wan to make totally new boards with these last modifications. My questions is about coils, 3uh and 22uh. How big they have to be? For how much "A" they have to be?
I have found 22uh/960ma and 3uf/1A. Will that be enough?
Few days ago i opened my quad405 which i have made some 10 or 11 years ago. I saw that one out of two 560ohm resistors on both amps are almost turned to coil! So i bought some 5w to replace those old 2w. Flipping the board to replace resistors i saw the horror! The horror was my work before 10 years 🙂 The board was some cheap thing, has really thin connections which are pealing very easy! I saw traces of my previous modifications. I think i was replacing output transistors.

So, to get to the point: I wan to make totally new boards with these last modifications. My questions is about coils, 3uh and 22uh. How big they have to be? For how much "A" they have to be?
I have found 22uh/960ma and 3uf/1A. Will that be enough?
I think these coils can be suitable:
http://uk.farnell.com/9754113/passives/product.us0?sku=epcos-b82111bc16
http://uk.farnell.com/516533/passives/product.us0?sku=epcos-b82144a2223k
http://uk.farnell.com/9754113/passives/product.us0?sku=epcos-b82111bc16
http://uk.farnell.com/516533/passives/product.us0?sku=epcos-b82144a2223k
Hi all!
Checking into this thread, as I'm very interested in these boards.
I'm modifying a couple of 405s at the moment, with very interesting results.
I'm just too scared to mess things up by modding the boards themselves though, but I am really curious about Keith Snook's various improvements -especially the C8 bridge cap.
P.S: Would anyone know off-hand if an AD845 could be dropped in straight in place of the TL071? I got the AD845 sitting here, and I have socketed the ICs, but I'm scared to fry the amp!
Regards
Patrik
Checking into this thread, as I'm very interested in these boards.
I'm modifying a couple of 405s at the moment, with very interesting results.
I'm just too scared to mess things up by modding the boards themselves though, but I am really curious about Keith Snook's various improvements -especially the C8 bridge cap.
P.S: Would anyone know off-hand if an AD845 could be dropped in straight in place of the TL071? I got the AD845 sitting here, and I have socketed the ICs, but I'm scared to fry the amp!
Regards
Patrik
AD845 are very sensitive devives that will easily oscillate. 0.1µF capacitors (or larger values) on the supply lines to ground is a must for most high speed devices.
Roger
Roger
RogerGustavsson said:AD845 are very sensitive devives that will easily oscillate. 0.1µF capacitors (or larger values) on the supply lines to ground is a must for most high speed devices.
Roger
OK, thanks for the advice!
I'll check on the scope with a cheap speaker before hooking up to my good ones.
BTW, the supply line cap would go from + to - on the OP's supply? (Or do I have to use two to ground?)
Edit: You specifically say "to ground" so I assume two are in order then.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Solid State
- Quad 405 DCD mod3 - PCB