Discussion in '"Boat Anchor" & Classic Equipment' started by WA9WFA, Oct 6, 2021.
Jim- Id like to hear as well! (Another) Scott ka9p
@WA9WFA and @KA9P,
First, I'd probably not use the HB-67 architecture (80 meter receiver with converter for other HF bands). But that's just me, and a subject for a different thread.
If I were to use the HB-67 architecture (which was also used in the Drake 2-A/B/C and the Mosley CM-1), here's how I'd fix the various issues:
5) The HB-67 uses a number of parts that were common in 1967 but will be hard to find and/or expensive today. The 7360 mixer tube, the two Compactrons, the various slug tuned coils, the audio iron, the crystals, the mechanical filter, the dial/coupling/tuning capacitor and bandswitch present some serious challenges if you don't have them already. There are work-arounds for some of these issues.
The 7360 issue can be dealt with two ways:
1) G4OEP has done impressive work with the 6JH8 as a receiving mixer. His circuit could be adapted to replace the 7360
http://g4oep.epizy.com/retro/retro.htm#The Retro Mk II
Lots of good info, both on the 6JH8 and other subjects such as audio-derived AGC and the 6BN6 as a product detector.
2) The Pullen mixer could be used instead. There is good info on it in on W7EKB's website. It's a simple circuit using a dual triode which reportedly gives low noise and high conversion transconductance.
and scroll down to "The Pullen Mixer". Several good articles, including some with practical circuit applications.
The Compactron issue can be dealt with by researching the characteristics of the sections and seeing if there exists a near-match in tubes that are more available, and redesigning the circuits to match. There are several triode-pentodes that would probably do the 6T9 audio amplifier job, or perhaps two completely separate tubes.
The slug tuned coils would have to be fabricated from whatever is available. Or, toroids and trimmer capacitors could be used for most applications. A Q meter and dip meter would be almost essential.
Crystals, filters and iron are where you find them. There have been 500 kHz filters on the watery auction place that might do the job. They're crystal filters, but appear to be OK.
6) The HB-67 covers 80/75, 40, 20, 15, and the first 600 kHz of 10 meters. No provision is made for full 10 meter coverage nor for the WARC bands (which of course were far in the future in 1967). These issues can be fixed but a larger chassis will be needed to accommodate the added parts, most of which are in the hard-to-find category - plus the bandswitch.
By using a larger chassis and a bandswitch with more positions, the WARC bands and all of 10 could be covered. The chassis could be drilled to accommodate the needed coils and crystals without actually wiring them until they became available. Also, it may be possible for the same coils to be used on adjacent bands (for example, the 10 meter coils might work on 12 meters).
With a 600 kHz tuning range, only three ranges are needed for all of 10 meters. This means coverage of 80/75, 60/WWV, 40, 30/WWV, 20, 17, 15, 12, and 10A/B/C would require an 11 position bandswitch. Some of the switching complexity could be taken over by small DIP relays (see below), reducing the number of poles needed on the bandswitch.
7) The HB-67 is optimized for SSB, and has some problems with CW and AM. The mechanical filter is far too wide for anything but very casual CW, while it is far too narrow for AM - and the BFO can't be turned off.
Additional filters could be provided for CW and AM, and switched with either a mechanical switch or small relays. There are some very small DIP packaged relays that are inexpensive.
8) In the HB-67, 40 meters tunes backwards (7.0 is at 4.1 MHz on the dial, 7.1 at 4.0 MHz, etc.) This is a minor annoyance if you're not used to it.
One solution is to use a 3400 kHz. heterodyne crystal for 40. This will put 7.0 MHz at 3.6 MHz on the dial, but the band will tune "forwards". 3300 kHz could also be used.
9) There is no 100 kc. calibrator, nor provision to adjust the tunable oscillator for heterodyne crystals that aren't dead on the marked frequencies.
Both can be added - another reason for a larger chassis.
10) The AGC is very basic audio-derived AGC, which is only applied to the two IF stages.
There are some very good AGC articles by W1DX in QST, and also the G4OEP source referenced above. Another reason for a larger chassis.
11) There is no gain control of the converter section; all stages operate at full gain all the time.
The converter RF amplifier could have its own RF gain control, or be on the RF gain line. Also AGC could be applied. Or, to be more adventurous, the RF stage could be eliminated and either a 6JH8 mixer or Pullen mixer used with no RF stage at all.
12) The tuning capacitor may not be the best choice for stability. Also, the unbuffered tunable oscillator isn't the most stable implementation.
The Miller capacitor shown has an aluminum frame and may have aluminum plates. There are better choices - the Polar capacitor used by some in the HBR-16 is excellent, but rare.
Another option is the tuning capacitor in the BC-455 ARC-5 receiver. It could be used as-is (with its own gear drive and dial), or, remove the gearing and use just the capacitor section. It is about 62 pF per section so some redesign of the tuned circuits would be needed.
The tunable oscillator section could be redesigned with a buffer as needed.
13) The chassis layout is generally good but doesn't allow much room for expansion, parts substitutions or changes.
Again, a bigger chassis. And an idea:
Use old chassis to make a chassis frame - just the sides and interior shield partitions, but no top. The top consists of various flat plates of aluminum, each with a section of circuitry, that are bolted to the frame. The sections could be: Converter, 80 meter tunable front end, IF, BFO, AGC, audio. Put the power supply on a separate chassis to free up space and reduce heat. That way different ideas could be tried easily.
Another approach is to split the receiver into modular sections. For example, one module would be the 80 meter receiver section only, with no power supply and the just headphone-level output. The converter section is completely separate, as is the power supply, and there's a speaker cabinet with audio amplifier built in.
14) There's no provision for muting the receiver when transmitting, which is odd, considering that the HB-65 predecessor had a muting jack.
Add muting jack. I use 1/8 inch phone jacks with the closed-circuit contact arranged so that the muting circuit is closed if no plug is in the jack.
15) The 600 kHz tuning range means that with a 36 turn dial the tuning rate is 16.6 kHz per turn. That's fast IMHO, although pretty common back-when. I'm not sure if the MD-8 dial is 36 turns for a half-revolution of the output shaft (actual reduction 72 to 1) or only 18 turns (actual reduction ratio 36 to 1).
This is a personal preference thing. I'm a CW op, and anything faster than about 10 kHz/turn is WAY too fast when using high selectivity. My gear typically has 5 kHz/turn or less.
The ARC-5 receiver capacitor mentioned above is about 30 turns to cover the range (60 to 1 reduction) so the rate would be about 20 kHz/turn. Might be OK for others.
16) The audio output is low impedance but the schematic says "PHONES". This may mean the overall gain isn't enough to drive a speaker - or - it may just mean they didn't provide switching to choose 'phones or speaker.
Build the audio section as a prototype and see how much gain and power output it has. Probably redesigned to avoid using a Compactron anyway.
A bit of work!
73 de Jim, N2EY
VE3CGA, Robert, here is a photo of the HB67 schematic. This is the first time I've tried to post a photo so we'll see if I did it right. Yes it does sound like it's the same receiver that you built. Yes the front end is double tuned using a configuration that I see quite a bit in the HB in the mid-late 60's. So you modified your version to replace the 7360 mixer with a dual gate Mosfet? I would be interested to hear your experiences building and using that receiver. Thanks! 73 Scott WA9WFA
Thanks Jim, I'm printing your recommendations out now and I'll go through them line by line and ponder. Just a couple of comments though. Yes some of the original parts would be a challenge to find such as the self contained Miller 1727 BFO assembly. I've never seen a this BFO assy before. Shouldn't be a problem though as 455 KHz BFO circuits with an adjustable transformer or coil are common enough to build. Yep 455 KC IF transformers have been a challenge also. Adjustable ceramic inductors have been difficult for me to find in the values that I need. For my current receiver project I had a Miller adjustable coil with quite a bit more inductance than the circuit needed, I had no other options at the moment, so I simply removed turns, tested with the LC meter, until the the inductance came into my range. It worked but was a bit risky depending on how stuck the wires were to each other as I was unravelling. If someone on here has some Miller adjustable ceramic inductors in their parts collection I'd like to get some. The Miller 1460 3 gang 20 pF variable may be the biggest challenge if I want to keep it original.
I don't want to deviate too much from the original design and building a hot rod receiver lest the project take me another 5 years to complete like my HBR-13. I'd like to start and finish in 2022 73, Scott WA9WFA
Jim said - "Use old chassis to make a chassis frame - just the sides and interior shield partitions, but no top. The top consists of various flat plates of aluminum, each with a section of circuitry, that are bolted to the frame. The sections could be: Converter, 80 meter tunable front end, IF, BFO, AGC, audio. Put the power supply on a separate chassis to free up space and reduce heat. That way different ideas could be tried easily."
Jim, I like this idea a lot, I used it while learning to do solid state stuff and then built a final non-modularized version of my main HB receiver and was very happy with the adaptability. But I fear it's a lot easier with small totally shielded boxes that are all 12V and 50 ohms in/out.
I could use your thoughts about proper power bussing in a modularized hollow state project. In my first frustrating endeavor building the Junior Miser's Dream, I ran into quite a few little, and a few big, issues with what I think I've learned is B+ impedance issues that reared their heads as motorboating as I changed configurations, added load, etc., and some hum issues related to power routing, and a bit of RF feedback, too.
If you were going to modularize an HF hollow state receiver within a large chassis with cutouts, how would you handled the power distribution issues? Let's say the PS is built on a separate chassis and provides B+, reg b+ and heater through a plug - then what? Would you try to just bundle it all to the back of the chassis and periodically bring out 4 pin connectors for various subchassis - or just wire modules up to a common point? And within the chassis would you separate the heater from the HV wires, or bundle it together, or even try to balance the heater lines rather than going single-ended? Is it necessary or overkill to supply the subchassis through HV feedthru caps (for example).
Happy for any thoughts you might have. After my last endeavor I swore I would follow a demonstrated successful layout, but retaining the experimental options is looking too tempting, and maybe necessary given parts availability, again
A most enjoyable discussion, and I'm learning much. Thanks, gents!
If you need some, I have....more than a few.
My solution has been to collect slug-tuned forms and then wind whatever coils are needed, checking inductance and Q on the Q meter.
Yes. I may have something you can use, but it would be larger.
"Fast, good, inexpensive - pick any two."
Here's an idea:
Redraw the HB-67 schematic into sections, and go over each section with the idea of "how build-able is this?" Then do as little change to it as possible, in order to match the parts available.
Then, build it on a larger chassis that was used originally, to allow room for future changes such as added IF filters, a buffered VFO, etc.
That way you can build a close copy, and then change it as you see fit in the future.
73 de Jim, N2EY
Regarding the two drives on the 1st receiver, I have no idea what that 2nd one is, it was in the junk box.
That receiver tunes very nicely, the 2nd one tunes a bit quicker.
I want to point out you do not need the worlds best mixer, no special tubes needed for very good performance.
All the things I have built were changed to support the parts I had on hand.
I think, these days, it would be very hard and take a very long time to get all the exact parts to build those old projects, even the simple ones.
Having built a few receivers, both tube and solid-state, I would say that the only part that could be dependent on the layout is the IF amplifier. It has sufficient gain to self-oscillate if laid out carelessly.
To avoid this, some care in the wiring layout should be exercised; keeping the stages in a straight line, use decoupling on heater, AGC and B+ lines, bypasses with short leads oriented in a "star" fashion around a common grounding point for each IF stage.
Otherwise, making the receiver in sections, each built in the Eddystone/Hammond die-cast boxes is a very tractable way.
A suitable division could be:
Front end, 7360 + first LO and mechanical filter in one box
IF, product detector, AGC and BFO in another
AF stage and power supply in a third
Each functional block can be tested and "debugged" separately, and the superior shielding properties of the boxes makes handling of any stability issues much easier.
If one is really shielding-conscious, a very good way of preserving the integrity is to make an internal partition in each box, with feedthrough filters for all supply voltages after they have entered the box through a suitable connector.
A further "point" in doing so is that all functional blocks become re-usable and can be used in other projects.
Some years ago, I built a few supplies in die-cast boxes combined with an audio amplifier using the ECL82. The supplies were dimensioned for supplying a "large" receiver or an exciter/small transmitter.
By digging deep in the "junk-box" a few transformers of unknown heritage were found which could provide unregulated voltages of around +600,+300 and -100 V, an OD3 provides 150 V regulated, two or more heater windings of 6.3 V 3 A each which can be either parallel or series connected to cater for various "flavours" of tubes.
The "down-sides" are that the boxes take up some space, and that some additional care has to be taken to avoid ground-loop induced hum problems.
Yes, you can also build sections on PCB boards.
That allows for changes without needing to re do a chassis, having extra holes, or running out of space.
Once the section works well, it can be added to the finished product.
At first, I laid parts out on the desk and tac soldered parts together to test LO designs.
I did not want tickler coils and other complex designs, I wanted one easy to make/get coil and an air variable cap
as the main parts.
Testing a few designs got me a good working LO in the ball park for range and output.