Palomar Regulated Bias Schematic Elite 450HD

[443]

Was reading over posts here about simple amplifier biasing to otherwise AM/FM only Class C amplifiers, by adding big resistors and reverse diodes, the normal, easy, but works, way to do it. So I thought I'd contribute this, for opinions, options, upgrades and applications.

This is a crude schematic I drew up from an amplifier I had here at my shop, it's an older, but later model, Palomar Elite 450HD (NOT the FET version). It originally came with 2SC2290A's x 4. The Bias is switched on at Transmit by an extra set of contacts on the TX/RX relay, a 4PDT being used as a 3PDT with two sets of contacts tied together for the TX Output. Not sure how "well" it regulated Bias Voltage or Current, but it did work well on SSB, lots of flowers. Circuit in this case used a L7805CV (ST Micro) and a TIP31C (EKL). Schematic may contain errors, but I tried to make sure I included everything. Sorry for the crudeness, just used MS Paint quickly to make it.

 

[2NC995]

Thanks for sharing, 443. I have a 4x2879 amp I want to add bias to. Been on the back burner for a while now.

 

[443]

Thanks for sharing, 443. I have a 4x2879 amp I want to add bias to. Been on the back burner for a while now.

You are welcome. I am hoping to receive some input from some of the more experienced amp techs here on this circuit before anyone tries it, just thought it might interest quite a few looking to do something other than the brick/sandbar resistor and reverse diode Bias thing. I would think this circuit if adequate could have a diode or another transistor added as a thermal tracker to further it's performance. Maybe someone can amend it and post their submissions.

 

[2NC995]

I gotcha. I’m not jumping off the cliff yet...haha.

Just idling along and learning.

I know that I want two things in my bias circuit....1, for it to be switched, and 2, for it to be regulated.

 

[Handy Andy]

Actually that is a good circuit to use, The Regulator takes care of the feeding - the TIP-31 takes care of the variability.

But remember - you're feeding 4 - this would work great for a simple 2 pill just as it is.

The problem comes up with "Current hogging".

Don't worry, the Diodes are mounted to wherever close to the pairs but - the issue is from the "Sharing" of the bias between the two pairs. One pair starts to heat up and will take most of the current - leaving the other pair with less to push itself into Class AB. This problem then can become catastrophic in that; you now have a pair of Transistors working harder than any other pairing

This can make the combiner - the one that puts the DC power sharing into both pairs (or how many you use) start acting weird and even burn up Loading is not equal.

If I may...

​

I've seen and would recommend that you "buffer" your branches off the TIP-31 output - so that you can share the loading EQUALLY amongst all the pairs.

It also acts like a "Break point" to help the technician find the pairs that go bad - the open ones raises the voltage present while the shorted takes all the current and voltage drops to zero at the RF Choke feeder for the pair.- just find the differences from the working sets.

The 100 ohms is arbitrary, you can go as deep into Class AB as you see fit by LOWERING the Resistor values but keep the precision so the loading is more equal.

• I see you're using diodes so that simplifies the circuit and when you have parings like this, remember that the Bias diodes themselves their VOLTAGE Drop will change. This was the concern I have to the equalized loading. The Resistor just limits power but not excessively.
• When a Transistor heats up, it will take in more power, so the need to keep the bias buffered helps the pairing so they don't "cascade" into more AB - the current starving the resistor provides to the pair, keeps this from happening - they can pull themselves out of AB and into Class C by starvation - less power curve in the cycle. Hence the cooling effect brought on by the Resistor pulling down the power available for the Bias they can use.

Just remember with each pair added, comes the cost of loading back at the TIP-31C. Each Branch will LOWER the effective Resistance TIP-31 "Sees" thru the Parallel - resistance in ohms' law So the current will go up - I'm just trying to protect the 7805 from the TIP-31 and dead shorts later in the feeder branches.

 

[Crusher]

I have always found it better to independently bias each two pill section. Then depending on circuit, you can thermal track each 2 pill/ transistor section independent of each other. Also just a heads up, bias voltage is a lot different on Toshiba transistors versus the new HG. Setting the HG at same voltage as a Toshiba and they will be turned on hard and run away/ go into oscillation. The HG only need about 0.475-0.5v. The goal is to bias so the pair of devices draws 200mA.

 

[443]

Thanks guys for the input. I find it useful and I am sure someone else here will too. Some of my thoughts:

Question about the diodes, the diodes in this circuit, are/were not mounted near the power transistors. I wasn't sure if their function was to kill AC/RF that gets into the Bias circuit, or ?? The arrangement looks to me like that's what it's for. If not please correct me. I wasn't thinking that they were used to track heat.

As for using for more than one pair of transistors. I see where you are going Andy, kinda like emitter ballast resistors in a power supply or audio amp with parallel transistors to even out the current/load the regulator transistor/supply sees.

Personally if I was to build an amp with this bias circuit, I'd use one circuit per pair of transistors and set them individually. I do a lot of SSB, so good biasing is a must.

 

[9C1Driver]

There is a good sticky note at the top of this section on this subject.

 

[443]

There is a good sticky note at the top of this section on this subject.

I read over that, but it was all about unregulated, sandbar resistor/reverse diode bias, which works "ok", but isn't regulated.

 

[443]

There is a good sticky note at the top of this section on this subject.

Also I did read over the Peakaboo thread (where is he these days? been a while since we spoke), but found lots of discussion, no schematics. I was intending this circuit as an alternate to the "Quick and Dirty Bias" sticky post. Being that this is a Regulated Bias design, "should" be better than the Sandbar resistor/Reverse Diode method. From reading over the Peakaboo post, I found this in a link, a LM723 with pass transistor based Bias circuit. The LM723 is a great regulator, but if I recall in past, having trouble with RF and that IC.. Maybe easy to overcome. The circuit is simple, parts R1 and D2 can be omitted for supply voltages lower than 40V (the max of the 723 IC). So for a 12V amplifier, no need for R1 and D2.

 

[443]

Though I don't want to get off topic and also go into information overload in this thread, just for reference, with the 723 regulator circuit above, this is the Dual Bias LNA schematic they were speaking of in the document.

 

[Crusher]

It’s been awhile, there is a decent bias design in Motorola Application Notes. Just don’t remember which one at moment. Bob85 and I discussed it somewhat years ago. If I get time in next few days. I’ll have to look it up.

 

[2NC995]

I read over that, but it was all about unregulated, sandbar resistor/reverse diode bias, which works "ok", but isn't regulated.

My initial plan was to use the "quick and dirty method", but with a 12v regulator.

Hopefully one day I'll get my plans off the ground...haha.

 

[Crusher]

443, your bias circuit after looking is very close to what I use. I suggest that 7805 regulator has some kinda of dropping/ current limit going to. If in mobile or on a power supply at roughly 14v. That’s 9v roughly the regulator has to drop. So yes, it better be heat sinked. I have opted to used 8v 3A regulator. Still use TIP33c or 3055. Motorola AN779 is bias circuit I started with. Just like design above but no regulator. The diode that gets grounded. I usually try to mount close or inbetween the pair of transistors right to heat sink. As heatsink heats up. Voltage will start to drop. In some of the old Kenwwod radios, they just put diode on top of transistor. Using a pair of 2290’s if I remember.

 

[bob85]

The 723 circuit 443 posted looks very much like the Motorola circuit using a diode rather than the base emitter junction of a transistor for thermal tracking,
works great.