Bremi BRL200 27-28Mhz Linear Amplifier

[crackle]

Last night I put the amp back together and gave it some power.

Results were encouraging, 100 watts out when fed with about 8 watts from the transceiver.
Quite sensitive on the tuning of the plate and load capacitors, but both peaked satisfactory. The HT voltage was 900 volts.
Output power seemed to fade slightly down to about 80 watts on a longer over.

Thanks for the information on the RF choke, I will have a go at making one and testing the amp with it to see if it has any effect on performance.

Mike

[G0HZU_JMR]

I think it is definitely worth trying to make a safety choke but obviously be careful when designing and fitting it. Get it wrong and the safety choke could overheat. Also, don't ever power the amp with the cover off even when testing the amplifier with the safety choke. You only have to make one casual or forgetful error once the cover is off. Amplifiers like this really should have a cover driven safety interlock because of the lethal voltages inside but this adds cost.

Back in the 1980s I refused to have anything to do with repairing valve CB amplifiers like the Bremi BRL200 because of the safety issues. I also didn't want any chance of any future blame of a shock or transformer failure being traceable back to me. I do recall visiting the flat of someone who had a Bremi BRL200 transformer failure. It was unpleasant to stay in the room for very long even a day after the event because the smell from the dead transformer was so strong

Output power seemed to fade slightly down to about 80 watts on a longer over.

I knew people who used these amplifiers for making DX contacts all afternoon every day and they can be quite rugged. I think the best thing is to only use them with SSB (low duty cycle) and to limit them to just under 100W PEP.

[G6Tanuki]

If you're trying to test the valve in these, the important parameter is peak emission - lots of Milliamps. The PL509 is rated to handle a peak cathode-current of 1.2 Amps (yes, honestly...) meaning you need quite a beefy tester. If it'll pass 500mA it's probably OK.

When the valve gets tired, the emission falls and so the output falls; people then turn the drive up - which results in 'flat-topping' on SSB signal-peaks and lots of distortion/splatter which will not make you popular with people operating on adjacent frequencies.

Given that these amps are generally designed to be used with antennas that present a DC-short across the output terminals, I wouldn't fret too much about the addition of 'safety chokes'.

[Techman]

Originally Posted by G6Tanuki

Given that these amps are generally designed to be used with antennas that present a DC-short across the output terminals, I wouldn't fret too much about the addition of 'safety chokes'.

Thank you, G6tanuki, that's exactly what I've been trying to explain and get across in a previous post

You may not be able to get a made up choke exactly right and it will waste precious RF and probably get warm in the process, and these amps need all the help they can get.

It's the same old argument regarding radios such as the Bush DAC90A and the live chassis and fitting an isolating transformer or modifying it with a captive mains lead, or the DAC90 with its asbestos surrounded dropper, or the old Fidelity record players with potentially live mains at the pickup cartridge. This is a vintage item made to Italian safety standards of the time (probably) and it is what it is!

There's no harm in experimenting with a choke if you want to and see how it performs. It looks like you're going to use it on 6 or 4 meters after all, so worth playing around with. My comment about not wasting time on it was regarding trying to put it back on 10-11 meters, which you'd struggle doing with half the capacitor vanes missing.

I've never heard of transformers burning out in the Bremi amps, it was the early Zetagi single valve models that burnt their transformers in abundance (wound for 220 volt mains, just), so I'm wondering if it was these that Jeremy was thinking about.

[Andrew B]

The old Zetagi BV131 amps with the single 6KD6 in them? The power meters used to melt and the valve used to "hole" if used enthusiastically, Both Bremi and Zetagi built amplifiers that used poor design and even poorer components.

[Techman]

Ah yes! I can't remember how much the Bremi amps cost new back in the day, but I do remember that the Zetagi BV131 retailed at £75 brand new, so if you compare that price to what a professional ham radio linear amplifier with all the extras would have cost at that time, then the question has to be asked as to what you would expect for your money.

You're right about the melting watt meter caused by heat from the valve, which I also think was partially responsible for the transformers burning out due to radiated heat. The heater voltage was something like 7.5 volts rather than 6.3 volts - these amplifiers had a short, but brilliant life! The later version with a single EL509/519 seemed to be a bit more reliable.

[Radio Wrangler]

There was a degree of optimism in the power rating of CB 'Burners' that has been seen in few other places.

David

[G0HZU_JMR]

I've never heard of transformers burning out in the Bremi amps, it was the early Zetagi single valve models that burnt their transformers in abundance (wound for 220 volt mains, just), so I'm wondering if it was these that Jeremy was thinking about.

It might have been, it was over 35 years ago. I was asked to repair various amp models like this numerous times but I always stayed away from them. However, the amp in the flat that had the smelly transformer was a Bremi BRL-200 because I got to see it myself.

Looking in an old ARRL handbook the various homebrew 1500W amplifiers all used a commercial safety choke made by Millen. A well made choke for mono band use ought to run cool at 100W even allowing for standing wave issues.

[crackle]

Originally Posted by Radio Wrangler

There was a degree of optimism in the power rating of CB 'Burners' that has been seen in few other places.
David

The power rating of CB type linear amps seems to be the figure that can be achieved in SSB mode. I believe it is normally almost double what would be achieved in AM or FM mode.

Anyway this amp is going to be put away and is not going to be used.

Reason; The short between the G2 and anode happened again, I believe it may be because the valve is operating in a horizontal attitude and the grids are also horizontal. The orientation of the valve is dictated by the valve socket which has the pins soldered onto the PCB. This may happen with any PL519.

But I did make a RF safety choke for the output, to help make the amp safer if it ever happened to be used with a long wire antenna.
The amp worked well with the choke when it was fitted loose above the mains transformer near to the output socket. But when fitted in the only available space vacated by the other valve I believe it was too close to the valve and its anode choke. The amp didn't like it and protested making odd noises and barely any output.

I was going to make a screening cover for the choke but that was when the G2 to Anode short happened again.

So anyway I have proved that these amps do work and work quite well for what they were designed to do.
They give a tiny bit of de-sense to the 2 CB'ers about 200m from my home on almost every channel of the CB band. But no real audible interference.
The re-design of the circuit from 2 x 6JB6 valves to a single PL519 or PL509 worked well, but was let down, I believe, because of the orientation of the valve.

I have finished the schematic diagrams.
The original schematic drawn by Rick Jackson of Euro Radio Co had many errors.
Here is the revised schematic with a clearer PSU layout, correctly drawn diodes and correct wiring of the Plate and Load capacitors.


This the schematic I have reverse engineered, showing the modifications made to use a singe PL519 valve.


Bremi_BRL200_corrected.pdf

WARNING
To any one else working on these amps please BE WARNED there is nearly 1000v DC floating around inside the chassis and under some conditions the capacitors can remain charged for hours after turning off.

Thanks everyone for all your help and advice.

Mike

[G0HZU_JMR]

Sorry to hear it didn't quite work out in the end, but at least the schematic looks a lot better now.

Not that I have any design experience with valve amplifiers like this but it's the anode choke that is the most challenging to design because it has to operate with such a huge RF voltage across it without overheating. The safety choke design should be easier to design because it should be fitted after the matching network. Ideally, it should be designed to survive operation into an open circuit where it might see a very high RF voltage from the matching network.

I googled for images of the BRL-200 to see the insides and it looks like the anode choke design is similar to the 50uH choke used in the big old HF grounded grid amplifiers made by Heathkit. These anode chokes were a single winding on some form of ceramic former. A while back I did have a go at winding the Heathkit version myself to see what resonances it had and what Rp it achieved over the HF bands.

The results for Rp were quite impressive apart from the 80m band. I wonder if the BRL-200 anode choke design was copied from a multiband choke as it seems to have a similar form factor with lots of turns of fine wire. If so it might have a similar inductance. Some manufacturers used cleverly spaced windings in the anode choke to achieve operation across all of the HF bands.

[G6Tanuki]

The design of pi-networks is something I studied in great detail a couple of decades back, when designing HF linear-amps for commercial/military applications. In particular, the anode-choke and its location.

I eventually settled on a design where the anode-choke was on the *output* side of the pi-network (so the DC component of the anode current passed through the pi-coil).

Why? Well, it meant that - under normal circumstances - the choke and HT-isolating capacitor were operating at the low-impedance end of the pi rather than the high-impedance end, meaning you could get away with a lot fewer choke-turns - which in turn meant there were less likely to be odd resonances-with-stray-capacitance, and two fewer directly-connected-to-the-anode-RF-hot components (the usual choke and coupling capacitor) so less opportunity for parasitics.

The pi coil-and-capacitors essentially act as a low-pass-filter to isolate any choke-self-resonance issues from the anode itself.

Also, it meant that the minimum-capacitance at the anode-end of the pi could be less - important when you're trying to cover 2 to 50MHz.

The only possible downside to this approach is that the pi-coil/tap-select switch and both the capacitors need to be able to handle the combined peak RF-voltage-swing-plus-HT-voltage, whereas in the 'classic' pi circuit they only have to handle the RF voltage.

[crackle]

Just re-visiting the circuit for the BRL200 again. In particular the voltage doubler circuit. Am I correct in thinking that R1 has no effect on discharging the capacitors, and it would be better to have a 470k resistor across each capacitor.

Mike

[Terry_VK5TM]

If you are referring to the circuit in the pdf of post #29, the 1M resistor will discharge the caps just fine.

However, if there is around 1000V across that resistor, I would change it to a series string of 3 x 330k resistors or 4 x 270k resistors.

[Terry_VK5TM]

There is just one other thing that has been niggling at me all day about the circuit which is:

The heater circuit arrangement, should it not be lifted above ground because of the 1000V HT to reduce the insulation stress on the transformer?

[crackle]

Originally Posted by Terry_VK5TM

If you are referring to the circuit in the pdf of post #29, the 1M resistor will discharge the caps just fine.

However, if there is around 1000V across that resistor, I would change it to a series string of 3 x 330k resistors or 4 x 270k resistors.

I had wondered if the single resistor would cope with the voltage.
Thanks

I have revisited this and got the valve working again.
I finished the screen for the output safety choke and can confirm with the choke in circuit connected to the output socket and ground, it is still merrily putting out 100 watts as measured on my MFJ 949E antenna matcher (set to through).

Here are some photos.
Those of a nervous disposition, please don't look at the last photo, which is how it was when it came to me.





Warning, never do this to a fuse.


Mike

[G6Tanuki]

Looking at the chassis, the big square hole under the valve... if you were to make a similar cutout in the bottom half of the case, do the same to the top case-half, put some expanded-metal mesh over both of them... it'd provide welcome cooling-relief to the valve.

[crackle]

If you notice someone has replaced the power meter with a VU meter. It fits very well and R4 has been changed to 7R so it gives a reading of +2 at 100 watts.
Mike

[Terry_VK5TM]

Originally Posted by crackle

Warning, never do this to a fuse.
Attachment 195969

Mike

And if either of those fuseholders are in the 1000V line, they are totally unsuitable and prone to flashover.

[G0HZU_JMR]

The suppressor resistor in the anode RL network looks like it has got a bit toasty at some point. This could be due to poor design of the suppressor causing the resistor to overheat with normal use. Or it could mean that the system is prone to instability. The suppressor resistor will often get toasted when an amp goes unstable at a higher frequency that is well out of band.

[crackle]

I hasten to add this was not my design, the 1M bleed resistor R1 and C3 in the single valve circuit are not actually as shown but are downstream of the fuse. So when the additional fuse in the HT circuit blows it disconnects the 1M bleed resistor and the capacitors can stay charged, for what seems like for ever. This was the reason I decided a few posts back to abandon the amp as it was so dangerous. I am thinking this fuse is a poor idea and it would be better to take the additional fuse in the HT circuit out. In the mean time I have added 4 x 220k resistors in series permanently across the HT capacitors.

Mike