|
Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment. |
|
Thread Tools |
11th Aug 2011, 10:18 pm | #21 | |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
Quote:
|
|
11th Aug 2011, 10:27 pm | #22 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Mullard 5-20 new chassis problems
By grounding the input cable twice you have introduced a small hum loop. Unlikely to be the cause of the current problem, so don't worry about it for now. However, as you seem to have introduced at least two ground faults I suggest you do some reading up on this when you have the amp showing some life.
|
11th Aug 2011, 10:38 pm | #23 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
Thanks Dave, I've removed the duplicate ground wire, but this did not affect the oscillation. Also installed the PIO caps and relocated that ground, to shorten and tidy wiring. Neither change has affected the oscillation. (BTW, the duplicate ground arrangement was part of the original chassis wiring, when the amp was completely silent in operation.)
I do struggle to understand good grounding principles, so any suggestions of good web references on this would be welcome. |
12th Aug 2011, 11:44 am | #24 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Mullard 5-20 new chassis problems
I don't offhand know any good web references for grounding, although I am sure there must be some. Better to learn the general principles:
- audio and RF grounding is different, as the former depends mainly on resistance and the latter mainly on inductance, so you can't always read across from one to the other. - keep signal ground and safety ground separate, apart from one good connection. The chassis should be safety ground. - anything which needs to be (signal) grounded should have one and only one ground connection, to eliminate loops. - high current loops (e.g. PSU transformer/rectifier/reservoir cap, output stage and its HT decoupling) should be wired as loops with one ground connection so the current does not flow through the ground connection. - minimise loop area, at least by routing wires together and where possible twisting them together. - regard all ground connections, however short and fat, as being a resistance (which they are) so think about what signal current will flow, what voltage will develop, and what will see and respond to that voltage. Bus or star grounding both work OK when done properly. A common mistake is to inject PSU charging pulses into the signal ground. I used a bus for my modified 5-20 amp. It was shared between channels, but the input stages and the output stages were at opposite ends, with the PSU currents kept well away. |
12th Aug 2011, 12:08 pm | #25 |
Dekatron
Join Date: May 2008
Location: Derby, UK.
Posts: 7,735
|
Re: Mullard 5-20 new chassis problems
Note that if you are using a ground bus, you should create a side branch for anything that might have hefty currents flowing through it. Otherwise, the potential difference inevitably created by said currents could affect the rest of the circuit upstream and downstream of the high-current section.
Code:
Wrong: -----##########----- Right: ########## # --------------------
__________________
If I have seen further than others, it is because I was standing on a pile of failed experiments. |
12th Aug 2011, 7:43 pm | #26 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
Thanks Dave and AJS. I've been thinking that I should revisit the physical layout recommended for the 5-20 and make sure that I follow it. The basic layout of my new chassis does do this with the large components, but I didn't follow the book exactly with the wiring layout.
Meanwhile, I'm familiarising myself with the controls on my Telequipment D67 scope. |
12th Aug 2011, 9:09 pm | #27 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Mullard 5-20 new chassis problems
The exact wiring layout is far more important than the physical location of components, although these two are obviously related.
|
12th Aug 2011, 9:32 pm | #28 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Mullard 5-20 new chassis problems
The EF86 voltages don't make sense. Check the earth connection of the 100 ohm resistor. That would fit with your observation that it stops oscillating but no audio with the NFB disconnected.
|
12th Aug 2011, 10:54 pm | #29 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
I agree about these voltages. I checked the earth connection on the 100ohm resistor. It was good, but I replaced it anyway, and also removed the long screened input cable to duplicate what I have on the other amp (following Mullard layout).
The voltages are no better, but higher: Anode 165V, Screen grid 192V and Cathode 9.6V. These are double what they should be. I'm still baffled. |
12th Aug 2011, 10:59 pm | #30 |
Heptode
Join Date: Jul 2009
Location: Stoke-on-Trent, Staffordshire, UK.
Posts: 719
|
Re: Mullard 5-20 new chassis problems
Can you give us a closer view of the EF86 holder and related tagstrip area, it might give a clue.
Pete |
12th Aug 2011, 11:00 pm | #31 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Mullard 5-20 new chassis problems
Disconnect the NFB and measure again.
The high cathode voltage could be rectification on grid 2 I guess. Last edited by PJL; 12th Aug 2011 at 11:10 pm. |
12th Aug 2011, 11:12 pm | #32 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
PJL: These measurements were taken with the NFB disconnected.
Pete: Will post a closer view of the EF86 holder tomorrow. It is not the same one as was there originally--maybe it has a physical fault? I've not thought of that. I guess I should also test with another EF86. |
12th Aug 2011, 11:49 pm | #33 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Mullard 5-20 new chassis problems
Measure the resistance from the cathode to earth. Then pull out the ECC83 and measure the voltages again.
|
13th Aug 2011, 12:07 am | #34 |
Hexode
Join Date: Jan 2011
Location: Aberdare, South Wales, UK
Posts: 403
|
Re: Mullard 5-20 new chassis problems
I think you may already have this but if not.........
http://www.r-type.org/static/5-20.htm I'm currently building a pair of my own. I think your problems are down to your layout. Its got to be neat and tidy. Looking at your pics....you definitely have a wiring error. Look at the busbar on the FOURTH tag, under the yellow cap C2. It should be on the SECOND tag, the ground end of the 100 ohm resistor R5.
__________________
Richard Last edited by trickie_dickie; 13th Aug 2011 at 12:14 am. |
13th Aug 2011, 12:09 am | #35 |
Heptode
Join Date: Feb 2006
Location: Roscommon, Ireland
Posts: 732
|
Re: Mullard 5-20 new chassis problems
Hi David
I know you said you checked it, but assuming the tag board layout follows the conventual 5-20 layout I cant help agreeing with Pete in post #16. The busbar should connect to the second tag from the right on the bottom row. In the photo it appears to connect to the fourth tag from the right. As all the voltages are high on the EF86 and the ECC83 I would check the bus bar connection to earth and check that the is no voltage appearing on the busbar where it connects to R5. I am probably way off the mark, it's just my 2p worth Best of luck with it Frank |
13th Aug 2011, 9:33 am | #36 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Mullard 5-20 new chassis problems
I think we all agree there is something wrong around the EF86.
If the 100 ohm and NFB were not connected there would be no path from cathode to earth so I would not expect to find 9.6V on the cathode. The busbar runs underneath the tagstrip and there is a wire just as it emerges that I suspect connects to the 100 ohm. If we take the ECC83 out then we should both prevent oscillation (and therefore the problem of incorrect readings from rectification) and isolate the EF86 from any higher HT source. Any fault can then easily be tracked down. |
13th Aug 2011, 11:19 am | #37 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
Thanks very much for these helpful pointers. Will follow up this afternoon, after attending to household duties.
Frank, I need to check if the tagboard follows the conventional layout exactly. This is the case with the other one I built, but I didn't build this amp and the only problems before were with the power supply layout. These I fixed with the guidance of a forum member in January. |
13th Aug 2011, 11:42 am | #38 | |
Heptode
Join Date: Feb 2006
Location: Roscommon, Ireland
Posts: 732
|
Re: Mullard 5-20 new chassis problems
Quote:
Frank |
|
13th Aug 2011, 12:20 pm | #39 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Mullard 5-20 new chassis problems
Frank, I don't think you are missing anything. A number of folks have picked up that the apparent cathode current does not match the apparent anode+screen current. You are assuming the circuit is otherwise wired correctly which does not look to be the case!
|
13th Aug 2011, 1:42 pm | #40 |
Octode
Join Date: Feb 2009
Location: Wallasey, UK.
Posts: 1,308
|
Re: Mullard 5-20 new chassis problems
You are right not to assume anything. I did go through it, marking every connection on the circuit diagram, but somehow one keeps on overlooking one's own mistakes.
Here are a couple of pics as it is now. |