Re: Are There any Electronic Design Engeering in this Group?
- From: "bg" <bg@xxxxxxxxxx>
- Date: Mon, 9 Jan 2006 20:39:19 -0800
Micro Henry wrote in message ...
>Electronic Design Engineer,
>
>I am Working on a Video Distribution Amplifier for my satellite system that
>uses 2N2219A as a common emmiter output driver. The unity gain bandwidth
is
>not high enough to control the phase shift properly. This results in a
>phase error in the the 3.58MHz color burst data. The error intensifies as
>the device gets warm and the Unit Loads(75 Ohms) i.e. UL, exceeds 4. It's
>supposed to handle 7. If no more the 4ULs it works fine.
>
>Now to the question. Do any of you video/broadcast engineers know of a
>substitute for the 2N2219A that has a higher Ft(unit gain Bandwidth). The
>2219 is about 300MHz with and Ic of 500ma, and can handle 1watt without a
>heat sink. It is a TO-39 can. So far, no one in MPLS has a heat sink, and
>But don't think that will entirely fix the problem. I believe it is a
>bandwidth problem. The hotter the device and the higher the load current,
>the lower the Hfe which results in a bandwidth/phase error problem. If I
>could increase the Hfe/Ft 25% or so I could move the operating point away
>from the rolloff point and minimized the phase error.
>
>This shows up as light blue -> light green, or Pink -> purple. Looking at
>the vectors on the color circle, this is a slight phase error in the color
>burst data.
>
>Looked on the internet for a replacement part with marginal success. Found
>nothing I wanted to try. If anyone has knowledge of a good substitute it
>would be greatly appreciated.
>
>uH
Obviously there are video buffer Ic's from National, Maxim etc.that would
make life alot easier. But --
An emitter follower can provide a frequency response that is nearly equal to
the transistors cut off frequency. This also implies very little phase shift
in the mid band. The color burst and color sidebands are far from a 2n2219's
300 Mhz Ft. I would suggest you measure the frequency response or phase
response to see what you actually have, or use a simulator like circuitmaker
which does both. If you are driving a single 75 ohm load, to 1 or 2 volts
peak, a collector bias current of 100ma is way to high. An Ic of 10ma might
be more typical. The source Z will have the greatest effect on the stage cut
off frequency. Even with a 75 ohm source, you should be able to get about a
100 Mhz bandwidth with negligible phase shift at 4Mhz. I would also suspect
that you might have an oscillation, or perhaps ringing due to a cable
mismatch if the color looks really bad. I've used 2219's before as video
cable drivers with no problem, and I've seen schematics that use similiar
output devices with lower Ft's.
.
- Follow-Ups:
- Re: Are There any Electronic Design Engeering in this Group?
- From: Micro Henry
- Re: Are There any Electronic Design Engeering in this Group?
- References:
- Are There any Electronic Design Engeering in this Group?
- From: Micro Henry
- Are There any Electronic Design Engeering in this Group?
- Prev by Date: Re: Newbie to CBand needs help
- Next by Date: PLEASE POST APPLE KEYNOTE IF YOU SEE IT!!!!
- Previous by thread: Are There any Electronic Design Engeering in this Group?
- Next by thread: Re: Are There any Electronic Design Engeering in this Group?
- Index(es):
Relevant Pages
|
