Midland 13-898B has LSB Only

[Justin B.]

I'm a sucker for vintage radios but they often come with "issues." I received a "new" FleaBay 898B today that the listing said will not power on, I figured it wouldn't be a big job. Got that part fixed which was the fuse on the PS board probably popped because electrical tape on connections to replaced bulbs had fallen off so probably shorted to chassis. Not sure why factories used solder-in fuses (maybe saved a nickle) so I swapped in a fuse holder. All of the bulbs were dead so I replaced them with 12V model train bulbs and continued on to the next issue.

Issue numbner 2 is that I seem to only have LSB, no AM or USB. Have good power output and modulation on LSB but zippo on USB or AM. Looking at the schematic I see that one set of four crystals, along with associated oscillator, covers LSB while another set of four, and oscillator, take care of AM and USB. That narrows it down pretty fine so tomorrow I'll get out the scope and start poking around. Maybe the bandswitch is still gunky or maybe the oscillator transistor Q102 or isolation diode D107?

73,

 

[nomadradio]

If you have a ham or shortwave receiver on hand, you can tune in one of the four AM/USB crystals and 'eavesdrop' on it to see if it's running. If you can hear the carrier from one of that set of four crystals, the gremlin is elsewhere.

73

 

[Justin B.]

Never thought of that. I'll have to dig out my old DX-440, it's the only "digital" SW receiver I have - all the rest are analog.

Thanks for the tip!

73,

 

[Justin B.]

Took my trusty old DX-440 out to the shop and performed the checks you suggested. On LSB I got 5 bars on the signal strength and on USB I barely got 1. So, the signal was present but extremely weak compared to LSB. Firing up the scope and poking around the LSB and USB oscillators I found the LSB oscillator "oscillating" but no waveform out of the USB osc. The voltages on Q106 look reasonable so I'm thinking that X115 might be dead. I'm heading back out to the shop to hook my signal generator, with proper 11.275 MHz freq, to the high-side of that crystal and see if the radio wakes up. I have used this method to test/isolate bad crystals in other radios - some old Lafayette tube-types come to mind...

More to come...

73,

 

[Justin B.]

After removing the sig gen from crystal I tweaked the trimmer for that crystal and it started working! I've had h\Heathkit HET Osc do that, too.

I think I jynxed myself wondering about what else the "tech" that twisted the wires together and taped on the replacement lamps might have jacked around with. When starting the alignments I found two crystals I couldn't get to come it - X102 and X103. When comparing the crystals to the schematic they were different! I subbed the sig gen onto X103 and they started working correctly. I'm wondering if some bonehead might have swapped these crystals in trying to get some "special" frequencies. I haven't did the math to see where the incorrect crystals would put the freq, maybe tomorrow.

Also, I found that if I run off 12V I can adjust the carrier balance for a reasonable null but still just a bit above spec. If I run off line voltage I have a hum/buzz and output is full. Maybe the buzz is getting into the mic circuit and causing the constant 12 W out. Have to look at that, too.

Took out the clock and found the motor frozen just like my other, was hoping to get a good clock out of the deal but no joy.

So, it looks like the hunt is on for a 7.7933 and 7.8100 MHz crystal.

73,

 

[Justin B.]

OK, putting my signal generator (1.5V signal) across the 11.295 MHz crystal the receiver came alive! I tested a few channels by testing it with a walkie and the reception was OK. I now have what I hope is one last problem to solve. In SSB I get full output on dead-key so the balanced modulator, or whatever this has, looks to be out of whack. I will probably need to go through all of the alignments just in case the same goober that spliced in the panel lamps decided to tweak everything!

One step at a time...

73,

 

[TM86]

One step at a time...

One the plus side, it looks like you've got this. Good work so far, looking forward to the conclusion.

 

[Justin B.]

After messing around with this I have made some determinations. A previous "goober" replaced X102 and X103 to get different channels. X103 was replaced with a 7.9100 MHz crystal which puts channels 9-12 at channels 36, 37, 38, and 40. X102 was replaced with a 7.9433 MHz crystal which will put it above the CB radio band, somewhere in 10M? I'll surf for it later. But, for now subbing in a DDS sig gen for X106, like I did on the other 898B, gives me all channels.

Issue number 2 is lucky half the transistors aren't blown up! Instead of 13.8V coming out of the PS board I have the full +22V. I am pulling out Q1 and Q2 for testing as Q1 has 22V on all legs... This is likely why I have full power out on SSB and why I can't null the carrier unless running on DC.

Back to the cave for further testing...

73,

 

[Justin B.]

OK, ordered a bunch of parts from Amazon.

1. I removed Q1 and found it "open" all the way around so subs for 2sC900 on order.

2. With the transistor out of the circuit I still get +22 V out so I found that the "pass transistor" Q5 2SD333 can have a TIP41 subbed so that's on order as well. I subbed in a 2n3904 and still got +22 V and it got pretty hot. I looked at the specs and the 2N3904 seemed to exceed the specs for the 2SC900 so maybe this is caused by number 3.

3. ZD401 is rated at 16V but I'm getting about +10V at the cathode end so possibly this is throwing off the transistor biasing for the whole regulation string. I added a zener assortment to my order.

4. The voltages around ZD2 and Q2 look reasonable so I think they are OK.

Parts should be here tomorrow so we shall see. If this circuit gives me much more grief I'm tempted to replace all of the reference related parts with a 7812 and be done with it!

Stay tuned...

73,

 

[Justin B.]

The parts showed up this evening and I took my goodies out to the shop. I replaced the 16V zener, Q1 with a 2N3904, and the pass transistor (Q5) with a TIP41. Firing it up everything was perfect and I was able to adjust the voltage to spec so I'm a happy camper. Q5 was completely shorted so that was probably why my original 2N3904 got so hot.

Tomorrow I'll wire a jack to run the VFO into like I did the other 898B and I have a few more parts on order to build another DDS VFO. I am going to use 1 2 line display which looks like it will fit perfectly in place of the old clock once I do some surgery on the front panel. But, that will be a story for a different thread!

Thanks for following along with my rambling.

73,