[EdHolland]

Not forgotten, but I have had other jobs to do, including fairly substantial car repairs.

This weekend I fabricated a better bracket for the IBAM board, and started to look again at the best layout for the pots, caps and resistors. Mounting can be achieved without drilling new holes, and the wiring change should be straightforward if I include the 220 K grid resistors on the bias board.

Cheers

Ed

[Ron Ramirez]

You'll get there on the IBAM.

[EdHolland]

Progress today, I figured out the picky perfboard pot placement problem, to get a layout and footprint of components that I like. It also makes sense in relation to the position of the output valves. Also, I chose carefully so that clockwise rotation of the trimmer will decrease the bias voltage, increasing the bias current. That makes most sense to me, turn the dial clockwise to get a bigger reading of the voltage on the cathode resistor.
Here's the board, ready to be cut down and wired in:

   

Wires to the o/p section will attach at the top of the board, for neat routing around the chassis. Feed from the -24V bias supply will attach to the left edge.

Am starting to feel much happier about how this will turn out. 

The board will get a dry run, just to see that the voltages are correct, and adjustable, before tying it in to the circuit proper.

Cheers

Ed

[EdHolland]

Lots of progress today and IBAM is working nicely.

I cut down the board, made mounting holes for the fixing bracket, and added wires for supply and distribution to each output valve. Then the board was tacked on to the -24v line, and the outputs all set to -18V - a good margin of safety for starting bias values.

The board was then fixed in to place, and I went about disconnecting the stock bias connections. This was achieved simply by detaching the 300 kOhm resistors from their tie in to the signal feeds. The loose ends were soldered back on to the tag for each resistor connected to the stock bias supply (connecting them to themselves). In this way, no parts were removed. Wires were run back to the -24 V point, and fly leads were connected from the IBAM board to the valve grids. I wanted to test proper, stable operation before the hard work of dressing all the leads along the chassis with the existing wiring.

Tests showed I had targeted the starting bias nice and low, with the voltage across the 10 Ohm cathode resistors indicating ~200 mA bias or a little more. Never having done this before, I tweaked away, then waited. The valves continued to settle in over several minutes, and I chased the adjustment to 320 mA. The multi turn pots are perhaps overkill, but weren't expensive, and the form factor suited construction on perf-board.

There was a bit of excitement when a probe slipped, and the screen supply was inadvertently connected to the cathode. The fusible resistor used in the cathode circuit did its thing... It was replaced (lucky the minimum order was 20), and I could not get bias to read sensibly on one of the valves. Alarm bells, large bias reading, a glance at the valve showed the anode just starting to acheive red heat. Power off! I'd disconnected one of the tacked flying bias leads when the "Crack!" of the short circuit made me jump. Fixed, checked, all well again.

With that, the four bias wires were routed carefully along the chassis, and the receiver tested again. Here is the finished installation. 

   

Cheers

Ed

[EdHolland]

Looking again at the bias voltages required to get uniform quiescent currents (anywhere from -16 V to -18.5 V) has made me think of investing in new output valves for this receiver. They are also something of a mixed bunch. Three are marked RCA, one Fisher, presumably the latter might be original. It certainly looks like it has seen better days - the Bakelite base is crumbling at the top edge from where the glass extends. They all look a bit well used. Sound wise, I have not judged the amp on a decent set of speakers yet - using either phones, or the cheap full range drivers on the workbench that are used for tests.

Since I plan on keeping this one, I looked around at the offerings for 7591a types, and also at reviews. There is a lot of concern for reliability with one brand, JJ. I can't confirm this, but they are certainly cheapest at ~$17 each. Better things are said of the Electro Harmonix brand, but they are packaged in a large envelope that doesn't look right, precludes addition of a wood cabinet, and puts everything even closer together. The last, probably best option is the Tung-Sol. Reviews are favourable (little report of trouble). They are quite expensive, at $30 apiece, but should last, now that bias is properly controlled. I have decided to go with the best option, and treat the Fisher to a secure future. These valves also look right, having the same envelope shape as the 60's types. I have the FM stage valves, but did need a few more 12AX7, so added those to the order.

Cheers,

Ed

[gregb]

Hey Ed, I have the new Tungsol's in a couple of different amps and I really like the way they sound. So far the reliability has been good and once bias is set they seem stable. Time will tell but so far a few hundred hours and I am very happy with them.

Gregb

[EdHolland]

That's good to know Greg - I look forward to a proper listening session of The Fisher.

[rfeenstra]

I have the TungSol's in my HH Scott with the tubes turned down a bit.  Sounds great, no problems. Several hundred hours.

[EdHolland]

I spent some time playing the unit today, listening to FM, and some CDs. Speakers were JBL bookshelf units. I am impressed, CD especially showed the performance of the amp, even limping along on the mixed bag of old valves. Sound was very good, and plenty of power, filling the garage with Jazz.

FM still needs a bit of attention. Every so often, I hear weird harmonic artefacts in stereo signals that vanish if the receiver is switched to mono. Looking at voltages, things are very high compared to the values quoted in the service manual... weak valves? I'll check when the new shipment arrives. I also need to complete the MPX alignment.

But... I can certainly see why this receiver has a following.

Cheers

Ed

[EdHolland]

I started through the FM alignment again today, focusing on what can be done on the multiplex decoder with limited equipment:

Signal generator, capable of audio through RF.
Fluke DMM
Communications receiver with 30 kHz lower end reception.

I do lack an oscilloscope, which is called for in the tests.

The first job is to measure bleed through of an 80 kHz signal, and set the coil, L100, to null it. I connected a wire to the test point, and wrapped this to an antenna wire (gimmick type coupling) connected to the receiver. 80 kHz was fed to the appropriate conection. It was easy to hear the rise and fall in signal, and set this filter per the instructions.

Next, 19 kHz, simulated FM stereo pilot tone is fed, and the transformer Z100 peaked up, measuring a DC voltage at the first section grid of V101. Straight forward and easy.

The next part is where a scope is called for. The 19 kHz is used to sync an oscillator running at 38 kHz, which operates the synchronous demodulation at V102. Normally this is done by establishing a stable 2:1 Lissajous figure on the scope. While tuning stations, I could use the receiver, tuned to the 38 kHz oscillator (actually a little off, to get a beat tone) and hear it lock in. No adjustment seemed necessary to Z101.

Returning to reception, things were improved a lot on all stations. The MPX has a high threshhold before auto switching to stereo (something I learned in an Audio Karma thread) so not all stations trigger the switch on my garage dipole. Those that do, are now crystal clear

Waiting for new valves now, and I'll call this one done

Cheers

Ed

[morzh]

>>Waiting for new valves now, and I'll call this one done

---Jolly good, old chap!

[EdHolland]

The consignment of empty state electronic devices arrived yesterday. No chance to test them yet - I was fixing something else - but the Tung-Sol valves/tubes/firebottles do look to be of good quality manufacture. I will be interested to see what it takes to bias them to the same current as the old ones.

Most of all though, I've been enjoying FM radio, while the unit is set up with some nice speakers in my fix-it space. I can see why these receivers continue to have a following.

[Phlogiston]

I have been working on some amps lately and have found 3 cases of "tested good" used tubes that have high impedance shorts, most likely deposits on the micas. So watch for any red-plating - might want to watch in dim light - watch for unusual hum and or excessive current across the OP trans. Run the tubes for a couple of minutes and then retest for shorts even a slight flickering of the short indicator. Add cathode fuses, if possible.

[EdHolland]

Good tips Russ - thank you

Odd instabilities might well show up from monitoring the current sense resistors put in place for the bias mod.

Besides, I like to watch/listen to tube gear with the lights out.

[Joeztech]

Ed;

Yes, the grid resistor's value originally chosen for the output tubes was marginal to begin with (actually a bit beyond the tube manufacturer's recommendations). The IBAM modifications to the negative bias circuit helps dial in the conduction in each output tube for the best possible mA balance in pairs of the output tubes. Adding 10 ohm 1/4 watt resistors in the cathode circuits of each output tube is part of the modification so that you can easily measure the current in each tube. Also consider adding 100 ohm resistors in series with the B+ to each output tube screen grid. Use 1/4 watt resistors there in order to serve as a fuse in case any  output tube shorts. This will help protect the valuable output transformers.

I have done one 500-C and two 800-C units so far. These units are great ones for reception in fringe areas. Use the Fisher alignment information developed for their KM-60 FM tuner kits. I recently compared the results of that method to using sweep alignment of two 800-C units and found that I could not improve on the results using the sweep alignment method!

Good luck!

Joe