I have got an old Kriesler radio gram 11-104 running after changing electrolytics and other capacitors etc. The radio has been running normally for several hours without any problems. Now to understand how and why! I have a reasonable 'textbook' knowlegde of how these radios work, but little practical experience.
The first valve is a hexode triode, 6AN7A. I have probed the DC conditions with a 20Mohm DMM, and a few voltage differences confuse me.
1. HT voltages are about 20 to 25 volts lower (225 vs 250 v) than circuit diagram. Is this any concern at this stage??
2. Grids 2 and 4 of hexode are at 78v compared with c-diagram of 80v (no signal). When radio is tuned to a signal, this voltage changes to 120/ 140 volts. Grids 2 and 4 are connected directly to Grid 2 of the following pentode double diode. Is this voltage increase normal?? Why??
3. Triode Plate voltage is 61v compared with 75v on c-diagram.
Triode Grid is -10.5v compared with -25v on c-diagram. Cathode at zero. This is particularly confusing as the 6AN7A data sheet indicates that a Triode voltage of around -10v is okay, whereas at -25v, the anode current would be practically zero. ???
4. Is a 6AN7A identical to a 6AN7?? Circuit diagram states 6AN7. I have a 6AN7A data sheet but cannot find one for 6AN7.

Any advice appreciated before I start checking individual components further.
Thanks, Ian

The usual explanation for low voltages is high resistors, however.......
you are using a digital multimeter.

Grids two are the screens and it is normal for a whole chain of them to be common. Normally they are also decoupled.

Having a circuit diagram to look at would also be useful. Somewhere on a good circuit, it will tell you what type of meter was used to obtain the figures. In most cases it wil be anologue, some times a VTVM.

Using a digital multimeter in the RF stages is a recipe for disaster. The high impedance meter will often give higher readings as it's circuit loading is negligable. The main problem is the RF; it will seriously impinge on most of those meters giving readings that are wholly misleading.

Many circuits will specify "no signal". If there is a signal then the AVC / AGC will start developing bias which is used to increase the negative voltage on the converters control grid, reducing its conduction.

25V low on the HT does sound a bit much, this will drag everything else low. The main cause is normally leaking caps. So I would hope that there are no paper caps left in it? If you changed resistors, the colours on some I have seen of late are quite dodgy and I have medical evidence (pilot & dyer) to prove I am not colour blind. Re check. Also look at the resistors being used as a HT choke, if they have gone that way?

If you can get an analogue meter of say 20,000 ohms per volt recheck using it.

Marc

Would making a removable 10k or 20k shunt block assist the digital in being more useful for voltage readings?

I would prefer analogue. The moments of inertia will render it insensitive to RF, and it will give the true DC.

With resistors it does not quite work like that:--

Analogue meters are expressed in terms of "Ohms per volt". This being at full scale deflection of the movement.

The meter draws current from the circuit and is therefore, a load on it.

Most of the old circuits quote values based on a meter of 1000 Ohms per volt. What this means is that for every volt the meter presents as 1000 ohm.

As this is taken at FSD, if the meter is on a 200 Volt range then it's 1000 x 200 = 200K . For the 10V range 1000 x 10 = 10K

If it was a 20,000 opv meter 20K x 200 = 4Meg: 20K x 10 = 200K

As you can imagine, this changes the whole loading on the particular circuit and is a constant source of apparent error.

All meters can produce errors, especially when used outside the parameters specified. All meters, used indescrimanantly (even digital) can change circuit conditions sufficiently to cause failures, especially in "Solid State" devices.

Marc

Thanks for the explanation Marcc, guess it would be useful to have an older meter on hand if to get serious about these things. My analogue is 10k per volt AC & 50k on DC..... never imagined poking around a valve radio when I got it however.........

The thing with DMM's is they are cheap and in many cases more robust; However you have now found one application where they have significant limitations.

Trying to set the mechanical regulators of Dynamo's (DC generators) is also another area where most DMM's "crash & burn", as the Dynamo output is "filthy".

This is why I could justify spending lots of dollars having the meter movement of an AVO 7X (500 opv) major overhauled.

I would get au fait with the anologue meter as they still have a place, especially if you are going to play with circuits with RF or AC riding on DC.

Marc

Marcc,
I have only started to appreciate this issue ie impact of test measurement equipment upon a circuit. I have an old analogue meter and will drag it out of storage and do some comparisons ..... after the Easter break.
I am sure that my good DMM states an input Impedance of 20 Meg ohms for DC volts. Circuit diagram states measurements require a 20,000 opv meter; then meter impedance required for 400 volts would be 400 x 20,000 = 8 Meg ohms. If correct, then my DMM should give a reasonable reading. Am I missing something???????
Ian

There are some Java voltage divider calculators floating around on the WEB, get one of these to play with. Often these are handy for finding the centre voltage on sets with screen dividers as well.

For the most part 8 meg does not present much of a load. However, when you consider a grid leak on a 6AV6 is typically 10Meg. you are close to halving it. This is also why leaking caps impinge significantly on grid bias.

The points that get into trouble are those where you test say a 200K on a plate with a 1000 opv meter on say 400V range 400K at 200V I would expect the indicated voltage to be half supply. (1:1 = divider). The open circuit voltage (no valve) with the 20K meter 195V. With lower ohmage resistors the 20K meter would be even less significant.

Marc

Thanks Marcc.
The grid leak that I am measuring across is only 47k. Analogue meter indicates about the same value, approx -10volts compared with the circuit diagram -25volts. The 6AN7A data sheet indicates that -10volts is well within a its working range whereas -25volts is close to cutting off the valve. Either the circuit diagram value of -25volts is a 'typo' or a 6AN7 is different from a 6AN7A valve or ????
???
Ian

Make sure that you have not got it on a station. As said that will force the bias up.

There are minimum about five ways to wire up a detector with AVC. If there is a resistor going to ground in the AVC and its, missing, high or dry; There is every chance that the voltage can go higher. Putting in new caps would exacerbate the problem.

If there is too much AGC that will reduce the sets ability to bring in stations..

Marc

Hi Marcc, Brad et al,
I haven't yet dragged out an old Analogue meter to do a comparison as discussed above, but maybe later this week.
However ..... I have identified the AGC and disconnected it from the 6AN7A. Screen voltage is now 112volts and independent of signal. AGC voltage (open circuit) is -0.5v (no signal) to -57volts (strongest signal).
As the circuit diagram shows, the screen resistor (27k) is in series with the screen; hence the changing voltage drop under different signal conditions. Interestingly, data sheets indicate a typical operating setup with a voltage divider supplying the screen voltage, ie two 27k resistors to provide half HT voltage, and obviously a more steady screen voltage.
Do you think that there is a good reason for this circuit design?
Any reason why I shouldn't experiment with a voltage divider circuit to the screen? (for educational purposes only)
Are the AGC open circuit voltages as you would expect in such a radio?
Thanks, Ian

The screen voltage on many converters is set in concrete.

All I would do there is make sure that the screen resistor is OK.

AGC voltages are not often specified due to the loading caused by the meter

Marc

You can find data for the 6AN7 at http://www.mif.pg.gda.pl/...084/6/6AN7A.pdf.

And more tube data can be found at http://www.tubedata.info/.

6AN7A is amost certainly a premium grade version of a 6AN7