HMV 653 Broadcaster all wave.

[ms660]

Originally Posted by sobell1980

Many thanks Tony and lawrence for your guidance on this. I will go check all the bias resistors shortly and report back. Other than experience, how do you know that V4 bias should be negative by several volts? or what it should be on any other valve for that matter? Just so i can learn and add it to part of my learning curve to aid my future diagnosis. I'm here to learn about vintage radio, not for a quick fix. I fully understand why valves are biased but not sure of what voltages to expect other than what is qoted in servie data. I know what to expect on the high and low tension side of the valves but not control or screen grids. I guess this is another new bit for me to learn on how the negative voltage bias is developed and where on valve circuits so it becomes routine for me to know what to measure and where. I've never experienced a fault really with resistors or biasing arrangements. If i could ask you to have the patience to talk me through it that would be great. I will also take the time to read my reference books, but, i find you can beat the knowledge and how things are explained by yourselves on here.

A lot is written about bias but to keep it simple in most valve situations a -ve control grid bias voltage is needed, and most valves in a radio will be operating in Class A excepting push pull output stages and the local oscillator that feeds the mixer.

The grid bias is -ve with respect to the cathode (that bit's important to understand) It sets the operating point of the valve on a chosen Vg/Ia plot, valve manufactures published valve data sheets/books, in those, apart from other data such as pin connections, max voltage/current ratings, heater supply etc are a set of characteristic plots (curves) Sometimes there are several sets but the main one to be concerned with so far as a basic understanding of grid bias goes are the Vg/Ia plots (grid voltage and anode current) these are derived and plotted by maintaining the anode voltage constant, changing the grid voltage (-ve bias) and recording the change in anode current for each change of grid voltage, this is done for different anode voltages, in other words each anode voltage will produce it's own plot for a change of grid voltage.

To illustrate this it's best to see a typical set of plots for different anode voltages, here's a set: http://www.r-type.org/pdfs/ecc82.pdf

Scroll down that page to the first set of graph plots (denoted A in the bottom right hand corner) As can be seen, the grid voltage is shown on the X axis (bottom) and the anode current in mA is shown on the Y axis (right hand side) the grid voltage is shown starting at the right hand side at 0 volts, follow that line upwards and it will intersect various anode voltages, these are the anode voltages mentioned earlier, they are prefixed with the letters Va, that means anode voltage, for instance Va170 means an anode voltage of 170 volts (with respect to cathode) So...with say 2.5 volts -ve grid volts and the vertical line from that point intersecting say the 170 volts anode voltage, turn right and you can now read off the anode current directly which is 14.5 mA, so if this valve was in a circuit and was in fully working order you would expect to record an anode current of 14.5 mA if the grid voltage was 2.5 volts -ve with respect to cathode and the anode voltage was 170 volts with respect to cathode, you can now see how this is fitting in with V4 of your receiver, if the grid bias voltage is known then the anode current can be arrived at by looking at the valve data curves, that value of anode current should more or less correspond to the anode current actually measured if the valve is functioning correctly.

So....let's do another one, why? Well 2.5 volts -ve on the grid gives an anode current of 14.5 mA at the said anode voltage, that current might be a bit to high for what we want so let's move to the left on the X axis to where the 5 volts -ve point is, this represents 5 volts -ve grid bias with respect to cathode (you'll remember that phrase from earlier on) Follow the vertical line at that point the until it intersects the anode voltage of 170 volts, turn right through 90 degrees, follow through to the anode current scale, you should read a corresponding anode current of 7.5 mA, a much lower current and well within the valves maximum ratings, let's call that grid bias/anode voltage intersection point the operating point....So.....There yer go, you've cracked that bit.

Ok so far but what about the signal input to the grid, well, two things need to be considered, the first are the valves characteristic plots, they all look fairly straight for a good part of their length except for the right hand one which is less so, for most single ended amplification purposes it's the "straight" part of the curve we are interested in, not the pronounced curved bit at the bottom, the point where the curve meets the bottom line is called the cut off point, you can see that it's dependent on the grid voltage for a given anode voltage, increasing the -ve grid voltage beyond that point will produce no change in anode current, the valve has cut off, in effect no anode current flows, reduce the -ve grid voltage from that point and anode current will start to flow again, so what does this mean for our signal?

The signal is superimposed on the grid bias voltage because the grid bias voltage is DC and the signal is AC, lets take our previous example, 5volts -ve bias on the grid and an AC signal is coupled to the grid, let's make the signal a 1 volt peak to peak sine wave, it's superimposed on the grid bias voltage of 5 volts -ve, this now means that the grid voltage will swing from 4.5 volts -ve through to 5.5 volts -ve, that will also produce a change in anode current which is equal for both halves of the sine wave either side of the bias voltage level, which is what we want, do a follow through and see, also note that the operating point is on the "straight" part of the curve.

If the operating point is shifted further down the plot onto the curved part by increasing the grid bias to 10 volts -ve then other things start to happen, distortion of the amplified sine wave starts, with the same 1 volt peak to peak input the grid voltage now swings from 9.5 volts -ve through to 10.5 volts -ve, if you follow that through you should see that the anode current flow for the +ve swing is greater than that of the -ve swing, not so good. It gets worse as the bias is increased, the operating point moves further down the curve, the negative half cycle at the anode gets smaller and smaller, and due to the increasing non linearity of the curve shown in the plot the signal gets increasingly distorted as a whole. When cut off is reached by increasing the bias even further the -ve half of the signal has disappeared altogether from any output at the anode and what's left of the +ve half will be very much distorted, do another follow through and see.

A plot picture is shown in this link and shows a sine wave in and out of a valve relative to the grid bias voltage on a Vg/Ia curve: http://www.r-type.org/articles/art-123a.htm

That picture should help you to clarify things that have been discussed, try printing out a few of the Vg/Ia plots of the valve under discussion and superimpose a sine wave of different amplitudes on them (input and output) at different bias points, it will give you a better feel for things.

Grid bias voltage can be obtained by various means such as a battery, the arrangement used in the receiver that's the subject of this thread, or via a cathode resistor where by the grid is coupled via a resistor to the -ve end of resistor that's connected in series with the cathode, as current flows through the cathode resistor the cathode will be +ve with respect to the grid, therefore the grid is -ve with respect to the cathode (which is what we want) Commonly referred to as cathode or automatic bias.

First internet port of call for valve data/characteristics, equivalents, pin outs etc is here: http://www.r-type.org/index.htm

This should help you with the basics, there's more to valves but as and when.

It woz late when I wrote this, crazy o'clock in the morning, any mistakes then guilty as charged.

Lawrence.

[ms660]

The resistors that form the bias potential divider chain are just about in the ball park except for R16, the resistor that develops the bias rail (R21) is spot on.

I would suggest replacing R16 then measure the -ve voltage across R21 again and all the tapping points of the potential divider, then to make sure that V4 isn't absolutely doing it's nut measure it's grid bias voltage again and then it's anode current by disconnecting the output transformers HT feed and connecting your meter set to 100 mA DC or whatever in series with that feed, It should give a better idea as to what's going on.

Lawrence.

[sobell1980]

Huge thankyous, Lawrence.
I'm just working my way through the valuable info you have provided me and using the links to understand it all. This is exactly what I was after. I'm just going over what you have posted several times, trying to remember it then going over the graphs in the links you supplied. I can now understand why v4 is so hot with no no bias. I've got an order with maplins for the bias resistors. I will get these replaced and try and come back to you with answers rather than questions this time. I have a better grasp now you have kindly taken the time to mentor me on this in great detail. I'm very thankful Lawrence. I will post dome results and findings over the next two days.
Dave.

[ms660]

No worries.

Keep at it.

Lawrence.

[kalee99]

Hello all,
Been following this thread from the beginning and its most interesting. I had one of these sets a few years ago. It was a "barn find", I seem to recall I had a similar fault which had me going round and around. The valve was getting red hot. I was lucky to have an Avo valve tester and suspecting the valve I tested it only to find it was 90% good & no inter electrode shorts cold or hot Then I got round to testing the valve holder, which on my set was those horrible thin Bakelite sandwich types, made for cheapness. Turns out their was a short circuit between two of the pin contacts down to the chassis because of the poor construction of the holder. Took me days to find this fault. Fitting a good ceramic octal base I had spare made all well and the set worked fine ever after.
From what I remember it was the anode pin shorted to one of the grid pins which in turn found a way to earth through a resistor? It was a long time ago but just though I would try to help.
Paul.

[sobell1980]

Well I got resistors R16 to R21 replaced last night and measured my v4 CG. Still no negative bias just a few hundred milli volts. I then checked my HT again at v4 anode and it was 50 v, most strange. I then checked my HT after L17 and yes it was 50 v here too. I then remembered that the CG voltage could be altered up and down before with the tone control. I tried turning the tone control and would you believe it the tone control is switching the HT on and off. Well between 50 and 300 v . I have two wires to the tone control. As you look at the back of the control I have a wire to the furthest left pin and one to the centre pin, nothing has ever been soldered to the furthest right pin.
Measuring across the the centre pin and left I get s max resistance of 280 ohms which decreases 17 ohms. When it reaches 17 ohms this is when it kills the HT. Measuring from the unused furthest right and then the furthest left pin I get the full quoted 1.5 m ohms which decreases to 280 ohms . The wire on the left of the tone control is wired to the CG of v4 and R16. I copied the wiring to how I originally found it but they may have been wrong as this set was such a mess. Something going on with this tone control and thus maybe also be killing the operation of the OPT too hence I have no speaked function at the minute either. Thought I would let you know my findings.
Thanks Paul for your advice. I have checked for shorts between all valve socket pins and to ground. Everything checks out ok fir now regarding that. Appreciate your input though.
Many thanks everyone.
Dave.

[ms660]

Sounds like the tone control might be defective, disconnect the two wires feeding it and connect a resistor between those two wires, I would suggest any value around 470k, then check out the voltages again.

Lawrence.

[ms660]

Ref my last post: Forgot to mention again, check V4's anode current as mentioned in an earlier thread.

Lawrence.

[peter_sol]

Have you replaced the tone control capacitor C24 on the trader sheet or C22 on your drawing?

[TonyDuell]

According to the trader sheet, only 2 connections of the tone control are used. One of the outside ones and the middle one. It does not matter which way round these are connected, it is a simple variable resistor.

One should go to control grid of V4, one end of R16, one end of C22.

The other should go to one end of C24 (there should be nothing else connected there)

The other end of C24 should go to the anode of V4

If everything is correct there is no DC path through the tone control. It sounds to me as though C24 is either incorrectly connected or shorted.

[sobell1980]

I've just gone over the tone control connections. I can't see anything a miss. One wire from the tone control goes to the CG valve pin where it also joins one side of R16 and one side of c22. The other wire from the tone control connects to C24 . The other side of C24 connects to the anode of v4 , the primary of the OPT transformer also connects to the anode of v4 . C22 and C24 have been replaced. I have also double checked the capacitor codes of c22 and C24 to make sure I installed the correct value.
I can find no shorts to ground in the v4 area. Just reading very high resistance readings to chassis with the set off so no direct shorts. Thankyou for your reply Tony. Appreciate your input on this. Tomorrow I will also try what Lawrence kindly suggests about substituting the tone pot for a fixed resistor, measure voltages again and also measure the anode current draw. Trying to remain positive and methodical and at least I'm learning all the time if nothing else at the moment. Many thanks for all help and replys.
Dave.

[TonyDuell]

The DC voltages (like the control grid bias) should be unaffected if the tone control and C24 are totally removed. So try disconnected them. If you can get the voltages right like that, you can then investigate the tone control.

[ms660]

If it still turns out that V4 is running very hot (as reported earlier) then things to consider are:

Insufficient control grid bias.

The valve has gone into grid current (grid current flowing) DC conditions for that would be a +ve voltage on the control grid, AC conditions for that would be an abnormally large grid input signal (remember those valve curves and the grid voltage swing due to the input signal to the grid, if it swings past the 0 volts bias point then the grid will go +ve and grid current will flow) This can happen if a stage has gone unstable and producing large oscillations, quite often in that case the oscillations can be out of the audio range eg: ultrasonic or higher, if suspected then a check with a 'scope will show what's what.

+ve DC on the control grid can also be caused by external leakage eg: coupling or feedback capacitors, valve sockets, valve bases, tag strips that are in the grid circuit that also have a +ve voltage on one of the other tags not in the grid circuit etc or internal leakage within the valve itself.

Lawrence.

[sobell1980]

Hi chaps. I've managed to grab 10 mins on a handful of occasions throughout the week between work and other jobs at home.
I removed both wires from the tone control and measured again. This time I get approx
-10 to -12 volts on the CG and the same either side of R16. This is with th tone pot disconnected. If i turn the volume up with the tone pot disconnected i get a high pitch squealing from V4 that is changeable with the volume control.
I then connected the highest value resistor i could lay my hands on, a 240k Ohm resistor across the two disconnected tone pot wires, this is to mimic the tone pot and eliminate it. I then measured again and had approx +2.00 volts then on the CG , the valve side of R16. On C23 side of R16 i still had approx -10 to -12 volts and no squealing from the valve regardless of volume position. I was hoping to see -7.00 volts on the control grid. The anode voltage is a steady approx 300volts and after a long period of switch on this does come down to 240 volts but we are talking several minutes. Still nothing from the speaker. Absolutely silent. I'm still unsure the OPT is correct with this extra secondary but i think this is a sepaerate issue in relation to the CG voltage and tone control.
Ive tested the tone control and get similar readings in resistance from both outer connections to centre connection. No reading or open circuit from any of the 3 tags of the tone control to its spindle or ground. However operation of the tone pot was 100% switching the HT on and off and the CG . Perhaps i need to look at my wiring previous to these stages and the other spurs taken from the resistor network.
Lawrence, do i remember you mentioning to disconnect something or not to to measure my v4 anode current? Perhaps disconnecting the spurs taken from between R17,R18 and R19 and R20 will change the readings for the better then i can acertain which part of the circuit is effecting this. Many thanks.
Dave.

[sobell1980]

Lawrence, just re-read your post about measuring anode current. Disconnect HT feed to anode from OPT and then connect my meter in series with this. Will try this also and get back to you,
Dave.

[sobell1980]

Continuing on from post #134 so members are able to follow on more easily with what tests and progress have been done today if but in dribs and drabs when time allows. I have tested the current draw on anode of V4 and it is 52 mA . This is with the tone control disconnected and 240k ohm resistor fitted in it's place across the the two tone control wires. This remains a constant and steady 52mA with the resistor in place regardless of volume control.
Anode current with the resistor removed is approx 42mA max volume and decreases and increases with volume control. Apart from goping over my wiring of previous stages I've kind of hit a brick wall at the moment. Just hoping if anyone can elaborate on my readings and findings. Many thanks,
Dave.

[ms660]

Anode current is too high, much too high, did you measure V4's control grid bias voltage?

Don't leave the receiver running for long at that anode current, seem to recollect in an earlier post that you have a 'scope, if so connect it to V4's control grid using an X10 probe, set the 'scope's Y channel to AC, set the scope's time base to 20 milliseconds, and see what's what on the grid, crank up the Y sensitivity if needs be, if there is a signal of whatever displayed then note it's amplitude, do this with no antenna connected, if a signal/waveform is displayed then note the effect of the volume control on it, a picture of the scope trace would be good if you are able to go down that route.

Lawrence.

[TonyDuell]

What anode current (V4) do you measure if the tone control is totally open circuit (i.e. the temporary resistor is removed)?

That resistor should not affect the control grid bias. Are you _sure_ C24 is not leaky? The fact that has been replaced does not mean the new part is good!

[sobell1980]

Tony,
With the temporary resistor moved i measure approx 42 mA anode current, This alters up and down with the volume control .

With the 240k ohm temporary resistor fitted it measures 52 mA . Adjusting the volume control does not alter this reading but it does with the temp resistor removed.

I will scope out the CG as Lawrence suggests. Hope i have a x10 probe!
I will also go over C24 again Tony. As you say its like it isnt blocking the DC and allowing the temp resistor and tone control to become part of a dc circuit switching the HT and bias up and down with either the fixed value resistor or variable tone pot.
Shall go do this now. Many thanks for taking time to reply.
Dave.

[TonyDuell]

42mA sounds too high to me. And I wonder how on earth it can vary with the setting
of the volume control.

More things to try : Leave the tone control disconnected (open circuit) for the moment/ The set will run without it.

Disconnect one end of C22 (Trader Sheet 732 reference). What anode current (V4) do you measure then? Does it vary with the volume setting?

Measure the voltages (with respect to the chassis) on the junctions of R17-R18, R18-R19. R19-R20, R20-Transformer centre tap. Do these voltages vary with the volume control setting?