BLF 278 amplifier

[Radio Garfield]

Hello, hope you are all right.

Friend of mine sent me faulty FM transmitter which have BLF278 amplifier. Exciter was faulty and after repairing it now I have to connect it to the amplifier. Before testing it with RF drive I checked the voltages. 47,3 V on Drain, Ugs = 1,35 V , Idq = 3,0 mA. Power meter show no self oscillation.

I have not any experience with blf278. I have built and worked a few blf177 amplifier before so I have few questions before firing it up.

1.) This Ugs of 1,35V and accordingly Idq of 3,0 mA seems to be bit low ? What do You guys think?

2.) Do You see any obvious problem on this board?

3.) What is the proper amount of RF drive, based on your experience, not datasheet? ( I saw that 1-2 W should be enough)

4.) What procedure you guys want to recommend me for starting this amplifier with RF input?
I see that people do it on two ways. First one would be to low BIAS on 0, bring a small amount of drive and bring BIAS up.
Second way would be to apply BIAS ( determine Idq first ), then apply RF drive from 0 to some value, watching the current and output
power.


I was reading some instructions from Albert H, here on this forum, that 10A fuse is good thing to have. This amplifier have 12.5 A fuse inside. I hope that it will work for safety.

Thanks guys.

P.S. We can comment about mechanical work and few other stuff later. It seems that this transmitter was on few places before it came to me

[RF-Head]

Thats the orginal amplifier that was mounted after your pll exciter
But whoww whats happend with the rest of the box ?????? also not the original Pll exciter is now in the box

[RF-Head]

it was like this one

Many of this type of transmitters are built and sold all over The Netherlands and the rest of the world
The one on Youtube is from a german buyer

[Radio Garfield]

All you said make sense. This guy that sent me all of this lives in Germany and while on new year vacation here in Bosnia, sent to me shipment with this FM transmitter and this exciter that is now used for driver of amplifier.

Well, those videos tell much really. Power supply is different but this in the box is also good. ( Mean-Well 600-48 )

But, since original exciter was faulty someone placed another one ( which is one that you see on the picture) and installed 12V power supply for new exciter. The person who was making these repairs did few bad jobs inside and that's way I am not sure in what condition is AMP-pallet also.

Since I don't have experience with BLF278, like I said before, I decided to ask here for guide-lines, before powering thing up.

Another problem was that, this complete box was not in use for long time and placed in some storage room with a moisture! I had to remove one of the input trimmer-capacitor . It was not able to move the plates anymore, welded somehow!
My questions from first post still stands please.

Thanks.

[Maximus]

The replacement exciter looks like an old version 8w tuneable RDVV.


Sent from my iPhone using Tapatalk

[Radio Garfield]

The replacement exciter looks like an old version 8w tuneable RDVV.

Sent from my iPhone using Tapatalk

[Analyser]

Hello, hope you are all right.

Friend of mine sent me faulty FM transmitter which have BLF278 amplifier. Exciter was faulty and after repairing it now I have to connect it to the amplifier. Before testing it with RF drive I checked the voltages. 47,3 V on Drain, Ugs = 1,35 V , Idq = 3,0 mA. Power meter show no self oscillation.

I have not any experience with blf278. I have built and worked a few blf177 amplifier before so I have few questions before firing it up.

1.) This Ugs of 1,35V and accordingly Idq of 3,0 mA seems to be bit low ? What do You guys think?

2.) Do You see any obvious problem on this board?

3.) What is the proper amount of RF drive, based on your experience, not datasheet? ( I saw that 1-2 W should be enough)

4.) What procedure you guys want to recommend me for starting this amplifier with RF input?
I see that people do it on two ways. First one would be to low BIAS on 0, bring a small amount of drive and bring BIAS up.
Second way would be to apply BIAS ( determine Idq first ), then apply RF drive from 0 to some value, watching the current and output
power.

The BLF278 is a quite a gainy device so doesn't need much drive. Give it about 3W input and that should be enough. You can then adjust the bias to increase the output power keeping an eye on the drain current, more that about 11A is too much.
You can periodically turn off the drive to see if the FET has started to draw any Idq, but you'll probably find that it won't need to.

You can always do it the other way round (like you said) and set, say, a 100mA bais current and then apply drive but you'll find the amp will probably end up less efficient because it'll need less drive.

When i said earlier that 3W input should be enough, that's for a well matched input and from my personal experience. You can increase to 4 or 5W but I wouldn't really go over 5W (especially with bias current being drawn).

[Radio Garfield]

@Analyser

Hello, thanks.

I have tested it tonight and I am worried that I have faulty MOSFET! I am not sure what else could be problem but I have no output power.
When adjusting the trimmers at input and that trimmer at output network I get to situation when about 15W goes out and that's about it!

I use two power meters. When I tune the input circuit of AMP drive power varies from 1-3 Watts. Output stays at same level of that 15W and don't varies at all.

I am confused. When I got this AMP, initial settings of BIAS were like this : 1,75 V on gate and only 3mA of drain current. I dont know is this any sign of faulty BLF278 or not? I remember on BLF177 I increase bias voltage until I have 5-10mA of drain current and then applying RF drive.

Can this same method be used on BLF 278? I mean, increase gate voltage until I have, lets say, 10-20mA?
IF BLF 177uses 5-10 mA , then BLF278 uses double value!

I am stuck now.

[Tec1 87.5fm London]

4 to 5w drive

[Radio Garfield]

Update:

Good news. MOSFET is not dead, power supply in the box is!

For some reason voltage drops down to about 20V and less when loaded. I noticed that green LED on power supply flickering and goes off completely when tuning. As soon as the load increase, voltage drop is higher. When I used my own power supply amplifier gave signs of life finally. I only had 41V but I can tune it to more than 60W of output, just to make sure that MOSFET is OK.
Phew, what a relief. Now I have to figure it out what is wrong with this Mean-Well.

[Radio Garfield]

4 to 5w drive

Thanks. I got Your reply while I was writing my own post.

[Albert H]

My old BLF 278 amplifiers used 4 Watts from a stage quite similar to the output of the PLL Pro III (2SC1947) and produced a nice, consistent 250 Watts with the power supplies I used. By cranking up the supply voltage (to about 60V), I found that the gain increased, so that I only needed about 2.5 Watts for full output. I used to have problems with the output transformer Teflon coax - I could get it to melt under some mismatch conditions! The amplifier was based on the Philips design notes, and was completely broadband (from 80 MHz to over 110 MHz), and had no adjustments - other than the bias.

I then discovered the Mitsubishi low voltage FETs and found out how to parallel them up. I've just done a PA with 8 of the 70 Watt jobs in parallel for 400 Watts out. The efficiency of these things is insane, and I only had to get rid of just over 120 W of heat. The only problem is the huge supply current, but I distribute so that each device has its own power input. The main supply comes in on a big piece of 16mm² twin & earth! The 8 FETs cost much less than Motorola "Gemini" device or a big Philips BLF....

[Radio Garfield]

My old BLF 278 amplifiers used 4 Watts from a stage quite similar to the output of the PLL Pro III (2SC1947) and produced a nice, consistent 250 Watts with the power supplies I used. By cranking up the supply voltage (to about 60V), I found that the gain increased, so that I only needed about 2.5 Watts for full output. I used to have problems with the output transformer Teflon coax - I could get it to melt under some mismatch conditions! The amplifier was based on the Philips design notes, and was completely broadband (from 80 MHz to over 110 MHz), and had no adjustments - other than the bias.

I then discovered the Mitsubishi low voltage FETs and found out how to parallel them up. I've just done a PA with 8 of the 70 Watt jobs in parallel for 400 Watts out. The efficiency of these things is insane, and I only had to get rid of just over 120 W of heat. The only problem is the huge supply current, but I distribute so that each device has its own power input. The main supply comes in on a big piece of 16mm² twin & earth! The 8 FETs cost much less than Motorola "Gemini" device or a big Philips BLF....

OK Albert. Thanks for info.

[radium98]

I then discovered the Mitsubishi low voltage FETs and found out how to parallel them up. I've just done a PA with 8 of the 70 Watt jobs in parallel for 400 Watts out. The efficiency of these things is insane, and I only had to get rid of just over 120 W of heat. The only problem is the huge supply current, but I distribute so that each device has its own power input. The main supply comes in on a big piece of 16mm² twin & earth! The 8 FETs cost much less than Motorola "Gemini" device or a big Philips BLF....

Albert H i would like to see a picture of 8 x 70watts in parallel with combiners or teflon transformers p;ease

[Analyser]

Albert H i would like to see a picture of 8 x 70watts in parallel with combiners or teflon transformers p;ease

So would I.

[Albert H]

The combiners are just inductors, emulating ¼-wave lines. The splitters are more hassle - they're small trifilar wound ferrites. They're fiddly to wind, but they work. The two PAs we built have left for the Middle East, but I expect to make some more for a client closer to home shortly. When we do the next ones, I'll make sure that I have some suitably anonymous photos to put up here.

The FETs are individually powered and their bias adjusted so that each stage develops 50 Watts - in fact they're done in pairs, for 100 Watts, but the principle's the same. There's a further benefit to using distributed power like this - you can "lose" a FET or two, without massive degradation of the output power - this was the principle used with the old BBC 300 Watt pallets with the pairs of TP9383s.

I find that the biggest problems at power beyond 150 Watts or so are the output filter capacitors. We've been making our own, out of brass and the Teflon™ insulation material. When they're trimmed to the exact value that we want, they're dipped in acrylic paint to stop them corroding. LB had the smart idea of painting stripes on to them denoting their values in pF! We used to just write on them with a Sharpie, but found that it could get wiped off quite easily.

The other problem is power supplies. I've never really been comfortable with running rigs from Switched-Mode Power Supplies. The problem comes when you want to do a really big linear supply - the transformers cost £100 and the smoothing capacitors are almost as expensive. You need heavy supply cables to prevent volt drop and heating issues..... I'm really going to have to have another look at the "Advance" and "Meanwell" supplies. These seem to be the best of a bad lot.....