Project: Re-engineered Murphy B40 Receiver

[radiozero]

I've no cabinet, or dial, but I have the parts from a Murphy B40C. So, I'm sort of going to re-engineer my own Rx from these parts I have. I will need to understand the circuit for this receiver, and I'm likely to make modifications. I'm hoping someone can explain the technicalities of the set and where improvements can be made. It will take a few months before I'm done.

Okay, let me start with the PSU. I've attached pictures of the two circuits of interest. Whether I will keep with these I don't know. From reports of those with Murphy B40D's there is no ac mains hum, so perhaps I should take the view, that if it's not broke, why try to fix it.

I'm likely to ask some questions.

Fig. 22 shows the power unit for B40B/C. Fig. 24 for B40D.

I know that C316 and C317 were added to reduce hum. Why they do, I don't know. Can anyone please explain?

Of course, I'll rely mostly on my own investigations, but every so often I might need some explanations. Rich

Oh bother, the pictures are non too good resolution!!! Aaargh!

[HamishBoxer]

Looking at where they are it seems to be to reduce AC hum?

David GM8JET

[radiozero]

Originally Posted by HamishBoxer

Looking at where they are it seems to be to reduce AC hum?

David GM8JET

C316 & C317 do reduce hum. But, technically why, I don't know. Someone will have an idea I think. :c)

C316 is 0.02 uF, C317 is 0.1uf. And in parallel.

[Leon Crampin]

These capacitors across the choke will cause it to become resonant at hum frequency, thus providing a high impedance path to the ripple current and reducing hum. It's a good trick if a receiver is to be used on only one supply frequency and Murphy used it quite a lot on some of their pre-war receivers. My A24 has it, for instance.

Note that the installed combined value of 120 nF will be resonant at 120 Hz (H.M. vessels used - and still use - a 60 Hz supply as part of a post war politics deal) so the values will need to be adjusted to be optimum at 100 Hz (50 Hz supply).

Leon.

[jimmc101]

I believe L304 is 20H and is parallel resonant with 0.12uF at close to 100Hz.
(103Hz theoretical but distributed capacity in the inductor is will reduce this.)

Thus the L304, C316, C317 combination will present a high impedance to the (100Hz) ripple component of the rectified waveform.

Jim

Leon beat me to it. The value of 20H came from here https://www.vintage-radio.net/forum/...&highlight=b40 post 6)

[radiozero]

Originally Posted by Leon Crampin

These capacitors across the choke will cause it to become resonant at hum frequency, thus providing a high impedance path to the ripple current and reducing hum. It's a good trick if a receiver is to be used on only one supply frequency and Murphy used it quite a lot on some of their pre-war receivers. My A24 has it, for instance.

Note that the installed combined value of 120 nF will be resonant at 120 Hz (H.M. vessels used - and still use - a 60 Hz supply as part of a post war politics deal) so the values will need to be adjusted to be optimum at 100 Hz (50 Hz supply).

Leon.

Hi. Well, I'd would never have figured that! Thanks a lot.

Also, those that are using B40's on 50Hz mains, are not getting the benefit of that hum reducing measure!!

[radiozero]

Just another thing: On Fig.22, would you include C306, 4 uF, (on the left hand side), as part of the smoothing capacitors in the PSU? Or, is it rather too distanced from the action, as it were?

I'm seeing C305, C314 & C315 as the smoothing capacitors, for sure.

[Alan Stepney]

A B40 "as is" works well.

With an additional trimmer to give fine tuning, it is even satisfactory resolving SSB on todays bands.
(In fact, it compares quite well with Yaesu and Icom Rx's.)

Hence, I would stick as close as possible to the original circuit.
Until (or unless) you find a specific problem.

[radiozero]

Okay, I've decided to make a few modifications as follows:

* PSU will be solid-state
* Local Oscillator will be solid-state

Rest will be as original i.e. hollow-state

PSU:

V302 (CV346, 7Z4) the full-wave rectifier will be replaced by two silicon diodes.

The salient characteristics for this valve are:

Maximum Peak Inverse Voltage: 1250V
Maximum DC Output Current: 100mA
Maximum Steady State Peak Plate Current: 300mA
Tube Voltage Drop (Tube Conducting 100mA each Plate): 40V

So, I've got to select a suitable rectifier, given the above info.

I will not require a 150V stabilised supply, so V304 (CV287) shall be removed.

Local Oscillator: .

This will be designed later. It will receive it's power from the 6.4V 0.9A winding on TR302, that becomes free due to replacing V302 with silicon diodes.

Any comments welcome.

[radiozero]

Okay, I'm not that up on component selection. This is what I've done so far: I gone to RS components. I've got to: Home>Semiconductors>Discretes>Rectifier. There are 1098 products. I've selected Peak Reverse Repetitive Voltage as 1200V. After that I've selected Maximum Continuous Forward Current of 0.5A (only current option available). I end up with component RGP02-12E-E3.

I am anywhere close to making a proper selection for the rectifiers? Thanks.

[murphyv310]

Be very careful of the peak voltages that will now occur with silicon rectifiers and possible long term component failure.
I understand that you don't have a full receiver and yes its fine to play about with the set pre se. I would not like to see a complete B40 modified in this way though as they are good performers and exceptionally tough!

[Herald1360]

With a 276-0-276 HT secondary, nominal piv would be 2*276*1.414=780V, so 1000V diodes should be fine. 1N4007 is cheap as chips so you could use two in series (total 4 diodes ~ 60p) if you feel happier. Put a 10n Xcap across each diode to kill any switching hash and a 100ohms or so in series with each leg of the rectifier.

[radiozero]

Okay, this is fundamentally the new circuit (without the Oscillator PSU that will feed from "y", the 6.4V 0.9A source). I've not included rectifiers in series or capacitors to kill switching hash, but know I can do that. Provides voltages of 200V, 230V and 250V.

[murphyv310]

Hi.
The 100 ohm in series with the diodes are still low resistor values in comparison to the cold load of the set, the initial HT will be much higher than the set would have been subjected to with a valve rectifier. The original rectifier was an indirectly heated cathode type and thus was a slow heating valve, so surge was never an issue.

The B40D uses two valves to reduce the voltage drop and reduce dissipation to improve reliability. My set is from 1954 and is still on its original CV valves. It has had a very hard life and still amazes! I wonder if the mod you are doing will last as long or make the set less reliable, I think I know the answer!

[ms660]

If you are solid stating the power supply the original caps in the IF deck and RF deck might moan at some point around switch on. If you are converting the local osc or bfo to solid state be aware the the original inductances for those circuits were carefully designed for valve loadings/characteristics etc, you might have to provide additionial compensation for drift if going solid state. Also if you want to keep the function of the calibration osc (extreemly usefull) I would consider leaving the BFO as is (a valve job) as the xtal in that set can sometimes be a bit fussy as to what it is in bed with, also on the bfo valve the screen volts are increased somewhat when switched to "cal" in order to drive the now xtal controlled osc well into harmonics in order to get up to 30 megs, this screen voltage is quite critical if using the original xtal, the screen will be fed via a 68k 1watt when in the "cal" position so make sure the ht at the other end of it is within spec on your psu conversion.
Hope some of this is of some use.

"No Fear, No Envy, No Meaness" (Clancy)

[ms660]

Mmmm me thinks similar Trev. Mines on every day usually tuned to Shanwick Oceanic while I'm fiddling with some other heap.

[radiozero]

I will be changing the original decoupling capacitors as well as the resistors. I wonder, does that mean I can get away with this hard-start PSU that I now have? If not, how would you engineer a "soft-start"?

[Skywave]

Originally Posted by Leon Crampin

Note that the installed combined value of 120 nF will be resonant at 120 Hz (H.M. vessels used - and still use - a 60 Hz supply as part of a post war politics deal) so the values will need to be adjusted to be optimum at 100 Hz (50 Hz supply).

Leon.

20 H and 0.12 µF resonate at 102.6 Hz. Temporarily ignoring the distributed capacitance of the inductor, a ±10% tolerance on either of these two components will produce a resonant freq. between the extremes of 93.3 Hz and 119.7 Hz.
Now making an allowance for that distributed capacity, the max. resonant freq. comes down to something like 100 Hz. For a typical loaded Q in the region of 5 to 10, the resultant bandwidth means that any adjustment for 100 Hz, if any, will be minimal.

Al.

[ms660]

Just suggestions.........1: increasing the value of the resevior somewhat will give you a few more millisecs but make sure the silcon diodes can take it and the fuses....2: knock up a series reg modified for soft start ie: 5 second sawtooth (simple exponential C/R circuit) to a shmit trigger/ thyristor/relay or something similar......3: keep the valve rectifier and take the supply for your solid state new build osc straight from the heater chain via bridge rectifier etc BUT NOT the heater winding thats feeding the rectifier as a heater cathode leakage or short in this valve would send your solid stae osc into orbit!..... 4: bolt an old 6 or 9 volt tranny to the decK and use that for all your LV requirements. Experience tells me that you should retain the original power supply and use suggestion 3 or 4 or keep everything as is if its there and working?
Hope this is of some use.

"I don't worry anymore I just think alot"

[glowinganode]

Or wire a (47K?) resistor between the rectifier and the res cap, with a bypass "Stand -by" switch.
Allow 30 seconds for the heaters to warm up before operating the switch.
Adjust the value of the resistor making sure the HT voltage at switch on doesn't exceed the normal operating HT voltage.
Rob.