30th Apr 2011, 12:23 pm | #21 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
Heres some more pics,the capacitor next to the blue F+T is not going to be used.And theres still work to cometicly,A
__________________
Curiosity hasn't killed this cat...so far. |
3rd May 2011, 10:02 am | #22 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
In an effort to calculate the value of the rectifier limiting resistors Ive found an equation,but am stumped as to what the star signifys,any help would be much appreciated.
The equation is Rt=Rs+(n^2)*Rp .Where Rs is the DC resitance of one half of the secondry.Rp is the DC resistance of the primary,and n is the ratio of the secondrys voltage devided by the voltage of the primary.Its the star/Asterisk-(star in original) thats puzzling me. In my last post I asked about the use of a 16/16uf capacitor in the Sterns plans,whereas in the original Mullard plans they use a 50/50uf cap.Any idea why? The rest of plan is identicle. Whilst on the subject of power supply,could I use a simple Villard circuit,unsmoothed, to boost the voltage of the 5v in the rectifier heater supply of the smaller of my two mains transformers.I think it should boost it by just enough (2v ish).Not sure what diode to use or if I,m happy doing so.Or should I just use a GZ34? I dont want to get a new mains transformer if I have to,Andy
__________________
Curiosity hasn't killed this cat...so far. Last edited by Diabolical Artificer; 3rd May 2011 at 10:06 am. Reason: Missing text/clarification. |
3rd May 2011, 10:31 am | #23 |
Retired Dormant Member
Join Date: Jan 2008
Location: Glasgow, UK.
Posts: 119
|
Re: Mullard 5-10 homebuild project
Hi Magnetic, A GZ34 heater takes 1.9A at 5V, An EZ91 1.0A, An EZ80, .6A.
I think either an 80 or 81 would work fine on the 5v winding, bearing in mind that the heater current is much less so the voltage would be slightly higher than 5v. EZ80 would be OK for the low loading version. The anode series resistors aren't critical, dont make them too high or they will run hot. Peter |
3rd May 2011, 10:50 am | #24 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Mullard 5-10 homebuild project
* is commonly used to denote multiplication, when only plain ASCII text is available. It avoids the potential confusion of using x, which could be one of the variables.
In this case you are taking the secondary resistance, and adding on the transformed primary resistance. |
4th May 2011, 2:11 pm | #25 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
Thanks again for your help.Regarding the problem of heater supply,Bob (Woodchips)has kindly offered me a 2 x 6.3v heater transformer.So thats sorted,the second EZ81 heater supply,and the heater supply for the EF86 preamp heaters.
As to limiting resistors,the Sterns plans call for two 300R resistors? ?According to my calculations(based on the valve Wizard site equation,and Daves info) I need two 33R for my bigger RS mains transformer (80+2.54^2x14= 170.3 ish-the 2.54 I got by dividing 334v by 245v,these being voltage at secondry divided by 245v,this being my supply voltage.)The EZ data sheet calls for 200R @300v. For my second smaller RS mains transformer ,which has a higher HT (657v opposed to 622V on the other MTF,I get (150+2.68^2x32=379.8,which is crazy! As the data sheet has a min of 240 ohms.(Impedance) Its driving me a bit nuts,as I want to get this right,it being my first homebuilt amp.And I still have to calculate the correct input impedance for a computor soundcard,a Akai R2R,and a CD player.My amp measures 330k resistance on these,but that not the same as impedance! Anyhoo its exersising the little grey matter,Andy.
__________________
Curiosity hasn't killed this cat...so far. |
4th May 2011, 3:06 pm | #26 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Mullard 5-10 homebuild project
80R sec plus 14R pri, with 334V sec from 245V pri gives 80 + 1.36^2 x 14 = 106. I'm not sure where you get 2.54 from. You need to add 100R to keep the rectifier happy. Maybe they add a bit more to drop the HT.
|
7th May 2011, 5:38 pm | #27 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
Got the two Brimar EZ81 valves through today,wired them up with two 300R 7w limiting resistors as per the Sterns data,and got 425v HT.That measured across a 50uf capacitor,but with no load.It should be 300v.Only left it on for 30seconds ish,just long enough to get a reading.I,m going to measure next with a BIG 120k resistor as load.Also have some 100R resistors on the way,as per Daves suggestion.Will get there in the end.I think I got my n-ratio wrong by using 620v in the maths instead of 310v.More pics to follow soon,Andy.
__________________
Curiosity hasn't killed this cat...so far. |
7th May 2011, 8:24 pm | #28 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,195
|
Re: Mullard 5-10 homebuild project
Hi Andy, as a rough approximation, the voltage will fall to about 70% of the value at no load, (when a large cap is used as smoothing/ resevoir) when loaded to the design current.
I would not expect 120K to have much effect on the voltage. A 10K, 10W resistor will draw 30 to 40ma and reduce the voltage to a level nearer to the operating point. Ed |
28th May 2011, 12:57 pm | #29 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
I,m just about getting to the end now,but am puzzled about a few things.In the plans there is an earth to to metal body of the amplifier,also a connection off pin 9 of the EF86.At the same time all metal body work is earthed to the earth lead.Is this correct procedure?
Lastly,I,m a bit at sea as to how the negative side of the HT works;it comes off the - terminal of the smoothing cap.No negative wire off the transformer.Does the - side of the HT return via the centre taps of the transformer?Just cant understand the physics etc here.Andy.
__________________
Curiosity hasn't killed this cat...so far. |
28th May 2011, 2:35 pm | #30 |
Dekatron
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,536
|
Re: Mullard 5-10 homebuild project
Negative HT...... well there has to be some connection somewhere between HT- and the transformer ct, though it may go via the chassis perhaps.
There is a good reason for taking the negative HT supply directly from the smoothing capacitor (and the positive for that matter). The reason is the ripple current through the capacitor. For minimum ripple voltage, the output DC has to be taken across the lowest impedance part of the smoothing circuit, which is the capacitor itself. Any extra wiring between the DC take off and the capacitor puts extra resistance in the ripple current path which means more ripple voltage on the DC. Ideal layout for a typical valve supply with centre tapped transformer and FW rectifier with resistive filtering would be: Twisted pair from TX outers to rectifier anodes. Single wire from rectifier cathode to reservoir capacitor +. Single wire from TX ct to reservoir capacitor -. Maybe a HT + take off for output valve from reservoir capacitor +. Resistor from reservoir capacitor + to smoothing capacitor +. Short wire from reservoir capacitor - to smoothing capacitor - (this may be a common connection if a dual capacitor is used, or chassis if can -ve types are used). HT + take off from smoothing capacitor +. HT - take off from smoothing capacitor - (or common terminal / chassis depending on arrangement). A lot of radios deviate from this ideal markedly but simply get away with it because of the inherent supply ripple rejection capabilities of tetrodes and pentodes and the relatively low resistance of the wiring compared with the ripple currents involved. It all gets much more important for low voltage high current supplies used in solid state gear.
__________________
....__________ ....|____||__|__\_____ .=.| _---\__|__|_---_|. .........O..Chris....O Last edited by Herald1360; 28th May 2011 at 2:41 pm. |
28th May 2011, 5:50 pm | #31 |
Retired Dormant Member
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
|
Re: Mullard 5-10 homebuild project
I'd 100% endorse that.
|
28th May 2011, 6:17 pm | #32 |
Administrator
Join Date: Dec 2002
Location: Cardiff
Posts: 9,073
|
Re: Mullard 5-10 homebuild project
From a cosmetic (and safety) point of view could you mount the output transformers through the chassis in the same way as the mains transformers? I assume they are held together with bolts so the flange on the contact side could be removed, the transformer lowered through the chassis, and the flange refitted underneath thus sandwiching the chassis. I don't know how that will work for transformer orientation relative to each other and the mains transformers though.
|
29th May 2011, 12:32 pm | #33 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
Thanks for those replys. And thanks Paul,I hadn't thought of mounting the OPT,s that way. Just about ready to pop the valves in.I thought I'd start with just the rectifier valve in, measure voltages etc and work my way up. Conected to a suitable dummy load that is.
I still have to sort the input side, I'm concentrating on getting the 5-10 up and running first. I'm not going to bother with the Mullard preamp. An update and pics to follow soon. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
31st May 2011, 1:57 pm | #34 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
Well we have lift off-ish.After a few false starts where I forgot to ground the - HT to chassis and left a resistor off the smoothing cap,I have a signal at the speaker.But it distorts badly when I turn the volume up a tad.At first I thought I,d taken the feedback off the wrong side,but after a few measurements,there is something iffy at the EF86.I,m measuring 47v @ the anode-it should be 60v,309v @g1,63v @ g2,308v @g3 and 307 @ the cathode.So something very wrong there.I,m getting 295v approx @ the anodes on the EL 84,s
Ive also had to add alot more resistance to the rectifier resisters,to bring down the HT voltage;it was about 20v,s too high.I had two 300r,s but have now got 580r each side,which gives me 309v at the first capacitor after the rectifier (smoothing) and 248v at the 2nd cap (resevoir). I,ll go back and check my wiring again to see where the fault is.The first EF I used had the catode shorted to the screen grids,maybe its done some damage?Anyroad,its the first valve amp Ive built and am quite amazed it didnt explode when I turned it on.Andy
__________________
Curiosity hasn't killed this cat...so far. Last edited by Diabolical Artificer; 31st May 2011 at 2:00 pm. Reason: Space at start of Paragraph lost? |
3rd Jun 2011, 3:42 pm | #35 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
In my efforts to try and trace the fault in this 5-10,I,ve hit a bit of a wall.The problem is that there is far too much voltage at g1,2,3 and the cathode of the EF86.When I disconnect the screen the situation improves.What I cant understand is how the screen works;on the plans pins 2 and 7 are wired together,and go to ground/- HT.How can the screen work if there is no current flowing through it? I just about understand how the grids work in a pentode,but am flumoxed with the operation of the screen in the EF86.The screen isn,t shown on the schematic.Andy.
__________________
Curiosity hasn't killed this cat...so far. |
3rd Jun 2011, 4:20 pm | #36 |
Nonode
Join Date: Dec 2002
Location: Coulsdon, Surrey, UK.
Posts: 2,135
|
Re: Mullard 5-10 homebuild project
Hi Andy
Welcome to the mysterious world of valves ! If I understand your post correctly, I wonder if you're confusing "screen grid" with screen ? Apologies if this is obvious, but something is clearly wrong with the connections to your EF86, or else the valve is faulty and drawing no current. Going clockwise around the EF86 base from underneath, pin 1> g2 or screen grid. To this should be connected a 470K ( to HT ) and a 47nF to cathode resistor ( 1k8 ) or direct to the cathode pin. Pin 2 > internal screen this should connect to pin 7 and both should then be earthed ( HT -ve strictly speaking ) pin 3 > (k)athode - should connect to pin 8 (g3 ) To both of these are connected the 1k8 cathode resistor. pins 4/5 heater pin 6 > Anode - to this should be connected the 120K anode resistor and also a connection to pin 2 ( or maybe 7 ) of the ECC83 Pins 7 & 8 - see above Pin 9 > is g1 - this should connect to the volume control (or output of previous preamp) ?) Ensure there is a path to ground via a 1Mohm resistor if the volume control is not here. The voltages should be around 50 - 100v on pins 1 and 6 and around 1-3 v on pin 3/8 The latter voltage should give an indication of whether the valve is conducting. If the voltages on pins 1 and 6 are almost the same as pins 1 and 6 of the ECC83, then the EF86 is not drawing any current. Hope this makes sense. I'm sure you'll find the problem quickly ! Andy |
8th Jun 2011, 11:21 am | #37 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
I,m getting there,but unfortunatly the amplifier distorts badly when I turn the volume up.Here are some measurements Ive taken. C1 314.6v,C2 306v,C3 252v,C4 131.5v.According to the measurement table in the Mullard book,theyre about right. EF86-a 51.5v,scr g 66v,k 1.524v - ECC83 a1 212.7v,g 50.89v, k 53v- EL84-a 299v, g1 ov ,g2 310v,k 10.61v-EL84 #2- a299v,g1 ov,g2 309v,k 9.5v.I,m 10v,s too low on the anodes of the EF and EL,s but 10v too high on the anodes of the ECC.Im wondering weather the problem lies with the ECC.
Heres a pic of an oscillogram at the output,input is 1khz @ 1v.Should I go through the amp with the osc probe from in to out to trace the fault? Having never done this am I right in attaching the ground lead of the scope to - HT and use the + part to test signal at the anodes and coupling caps?Your guidance as always much appreciated,Andy
__________________
Curiosity hasn't killed this cat...so far. |
8th Jun 2011, 12:55 pm | #38 |
Dekatron
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,536
|
Re: Mullard 5-10 homebuild project
Just beware that the dc voltage rating of the ac input option on the 'scope is high enough.
The trace shown seems to have some hum on it, as well as looking awful if it's supposed to be a sinewave. Are the drives at the output valve grids clean, of equal magnitude and opposite phase?
__________________
....__________ ....|____||__|__\_____ .=.| _---\__|__|_---_|. .........O..Chris....O |
8th Jun 2011, 4:10 pm | #39 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,669
|
Re: Mullard 5-10 homebuild project
Thanks,I,ll check that.Ive been checking all the wiring again and I cant find anything amiss.I tried two different EL84,s-same result.Here are some more scope pics;sorry about the quality.#1 is @ output,no input.#2 is taken from the two coupling caps to EL grids,from the ECC phase splitter. (pin 2).#3 is output- 100hz sine input,#4 is 100hz input sine.I,ll do what you suggest Herald.thanks for your help.Andy.
__________________
Curiosity hasn't killed this cat...so far. |
8th Jun 2011, 11:38 pm | #40 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Mullard 5-10 homebuild project
That looks like clipping. Your sine wave is being turned into a square wave, but then there is some LF droop at the output. Either the coupling caps to the output grids are too small, or the grid leak resistor is too small, or the output transformer is not big enough. Is the feedback loop closed or open? What is the voltage of the sharp rise (or fall) in the output?
The phase splitter is DC-coupled to the first stage, so the slightly low voltage at the EF86 anode causes the slightly high voltage at the ECC83 anodes. This may mean that the ECC83 is biased too near cutoff. |